Unnamed: 0
int64 0
5k
| title
stringlengths 9
210
| abstract
stringlengths 164
1.92k
|
|---|---|---|
4,000 | Federated Self-supervised Learning for Heterogeneous Clients | Federated Learning has become an important learning paradigm due to its
privacy and computational benefits. As the field advances, two key challenges
that still remain to be addressed are: (1) system heterogeneity - variability
in the compute and/or data resources present on each client, and (2) lack of
labeled data in certain federated settings. Several recent developments have
tried to overcome these challenges independently. In this work, we propose a
unified and systematic framework, \emph{Heterogeneous Self-supervised Federated
Learning} (Hetero-SSFL) for enabling self-supervised learning with federation
on heterogeneous clients. The proposed framework allows collaborative
representation learning across all the clients without imposing architectural
constraints or requiring presence of labeled data. The key idea in Hetero-SSFL
is to let each client train its unique self-supervised model and enable the
joint learning across clients by aligning the lower dimensional representations
on a common dataset. The entire training procedure could be viewed as self and
peer-supervised as both the local training and the alignment procedures do not
require presence of any labeled data. As in conventional self-supervised
learning, the obtained client models are task independent and can be used for
varied end-tasks. We provide a convergence guarantee of the proposed framework
for non-convex objectives in heterogeneous settings and also empirically
demonstrate that our proposed approach outperforms the state of the art methods
by a significant margin. |
4,001 | SS-GNN: A Simple-Structured Graph Neural Network for Affinity Prediction | Efficient and effective drug-target binding affinity (DTBA) prediction is a
challenging task due to the limited computational resources in practical
applications and is a crucial basis for drug screening. Inspired by the good
representation ability of graph neural networks (GNNs), we propose a
simple-structured GNN model named SS-GNN to accurately predict DTBA. By
constructing a single undirected graph based on a distance threshold to
represent protein-ligand interactions, the scale of the graph data is greatly
reduced. Moreover, ignoring covalent bonds in the protein further reduces the
computational cost of the model. The GNN-MLP module takes the latent feature
extraction of atoms and edges in the graph as two mutually independent
processes. We also develop an edge-based atom-pair feature aggregation method
to represent complex interactions and a graph pooling-based method to predict
the binding affinity of the complex. We achieve state-of-the-art prediction
performance using a simple model (with only 0.6M parameters) without
introducing complicated geometric feature descriptions. SS-GNN achieves
Pearson's Rp=0.853 on the PDBbind v2016 core set, outperforming
state-of-the-art GNN-based methods by 5.2%. Moreover, the simplified model
structure and concise data processing procedure improve the prediction
efficiency of the model. For a typical protein-ligand complex, affinity
prediction takes only 0.2 ms. All codes are freely accessible at
https://github.com/xianyuco/SS-GNN. |
4,002 | A Convergence Theory for Over-parameterized Variational Quantum Eigensolvers | The Variational Quantum Eigensolver (VQE) is a promising candidate for
quantum applications on near-term Noisy Intermediate-Scale Quantum (NISQ)
computers. Despite a lot of empirical studies and recent progress in
theoretical understanding of VQE's optimization landscape, the convergence for
optimizing VQE is far less understood. We provide the first rigorous analysis
of the convergence of VQEs in the over-parameterization regime. By connecting
the training dynamics with the Riemannian Gradient Flow on the unit-sphere, we
establish a threshold on the sufficient number of parameters for efficient
convergence, which depends polynomially on the system dimension and the
spectral ratio, a property of the problem Hamiltonian, and could be resilient
to gradient noise to some extent. We further illustrate that this
overparameterization threshold could be vastly reduced for specific VQE
instances by establishing an ansatz-dependent threshold paralleling our main
result. We showcase that our ansatz-dependent threshold could serve as a proxy
of the trainability of different VQE ansatzes without performing empirical
experiments, which hence leads to a principled way of evaluating ansatz design.
Finally, we conclude with a comprehensive empirical study that supports our
theoretical findings. |
4,003 | FBNETGEN: Task-aware GNN-based fMRI Analysis via Functional Brain Network Generation | Functional magnetic resonance imaging (fMRI) is one of the most common
imaging modalities to investigate brain functions. Recent studies in
neuroscience stress the great potential of functional brain networks
constructed from fMRI data for clinical predictions. Traditional functional
brain networks, however, are noisy and unaware of downstream prediction tasks,
while also incompatible with the deep graph neural network (GNN) models. In
order to fully unleash the power of GNNs in network-based fMRI analysis, we
develop FBNETGEN, a task-aware and interpretable fMRI analysis framework via
deep brain network generation. In particular, we formulate (1) prominent region
of interest (ROI) features extraction, (2) brain networks generation, and (3)
clinical predictions with GNNs, in an end-to-end trainable model under the
guidance of particular prediction tasks. Along with the process, the key novel
component is the graph generator which learns to transform raw time-series
features into task-oriented brain networks. Our learnable graphs also provide
unique interpretations by highlighting prediction-related brain regions.
Comprehensive experiments on two datasets, i.e., the recently released and
currently largest publicly available fMRI dataset Adolescent Brain Cognitive
Development (ABCD), and the widely-used fMRI dataset PNC, prove the superior
effectiveness and interpretability of FBNETGEN. The implementation is available
at https://github.com/Wayfear/FBNETGEN. |
4,004 | sat2pc: Estimating Point Cloud of Building Roofs from 2D Satellite Images | Three-dimensional (3D) urban models have gained interest because of their
applications in many use-cases such as urban planning and virtual reality.
However, generating these 3D representations requires LiDAR data, which are not
always readily available. Thus, the applicability of automated 3D model
generation algorithms is limited to a few locations. In this paper, we propose
sat2pc, a deep learning architecture that predicts the point cloud of a
building roof from a single 2D satellite image. Our architecture combines
Chamfer distance and EMD loss, resulting in better 2D to 3D performance. We
extensively evaluate our model and perform ablation studies on a building roof
dataset. Our results show that sat2pc was able to outperform existing baselines
by at least 18.6%. Further, we show that the predicted point cloud captures
more detail and geometric characteristics than other baselines. |
4,005 | Augmentation-induced Consistency Regularization for Classification | Deep neural networks have become popular in many supervised learning tasks,
but they may suffer from overfitting when the training dataset is limited. To
mitigate this, many researchers use data augmentation, which is a widely used
and effective method for increasing the variety of datasets. However, the
randomness introduced by data augmentation causes inevitable inconsistency
between training and inference, which leads to poor improvement. In this paper,
we propose a consistency regularization framework based on data augmentation,
called CR-Aug, which forces the output distributions of different sub models
generated by data augmentation to be consistent with each other. Specifically,
CR-Aug evaluates the discrepancy between the output distributions of two
augmented versions of each sample, and it utilizes a stop-gradient operation to
minimize the consistency loss. We implement CR-Aug to image and audio
classification tasks and conduct extensive experiments to verify its
effectiveness in improving the generalization ability of classifiers. Our
CR-Aug framework is ready-to-use, it can be easily adapted to many
state-of-the-art network architectures. Our empirical results show that CR-Aug
outperforms baseline methods by a significant margin. |
4,006 | Linear Algorithms for Nonparametric Multiclass Probability Estimation | Multiclass probability estimation is the problem of estimating conditional
probabilities of a data point belonging to a class given its covariate
information. It has broad applications in statistical analysis and data
science. Recently a class of weighted Support Vector Machines (wSVMs) has been
developed to estimate class probabilities through ensemble learning for
$K$-class problems (Wu, Zhang and Liu, 2010; Wang, Zhang and Wu, 2019), where
$K$ is the number of classes. The estimators are robust and achieve high
accuracy for probability estimation, but their learning is implemented through
pairwise coupling, which demands polynomial time in $K$. In this paper, we
propose two new learning schemes, the baseline learning and the One-vs-All
(OVA) learning, to further improve wSVMs in terms of computational efficiency
and estimation accuracy. In particular, the baseline learning has optimal
computational complexity in the sense that it is linear in $K$. Though not
being most efficient in computation, the OVA offers the best estimation
accuracy among all the procedures under comparison. The resulting estimators
are distribution-free and shown to be consistent. We further conduct extensive
numerical experiments to demonstrate finite sample performance. |
4,007 | Investigating Information Inconsistency in Multilingual Open-Domain Question Answering | Retrieval based open-domain QA systems use retrieved documents and
answer-span selection over retrieved documents to find best-answer candidates.
We hypothesize that multilingual Question Answering (QA) systems are prone to
information inconsistency when it comes to documents written in different
languages, because these documents tend to provide a model with varying
information about the same topic. To understand the effects of the biased
availability of information and cultural influence, we analyze the behavior of
multilingual open-domain question answering models with a focus on retrieval
bias. We analyze if different retriever models present different passages given
the same question in different languages on TyDi QA and XOR-TyDi QA, two
multilingualQA datasets. We speculate that the content differences in documents
across languages might reflect cultural divergences and/or social biases. |
4,008 | Recipe for a General, Powerful, Scalable Graph Transformer | We propose a recipe on how to build a general, powerful, scalable (GPS) graph
Transformer with linear complexity and state-of-the-art results on a diverse
set of benchmarks. Graph Transformers (GTs) have gained popularity in the field
of graph representation learning with a variety of recent publications but they
lack a common foundation about what constitutes a good positional or structural
encoding, and what differentiates them. In this paper, we summarize the
different types of encodings with a clearer definition and categorize them as
being $\textit{local}$, $\textit{global}$ or $\textit{relative}$. Further, GTs
remain constrained to small graphs with few hundred nodes, and we propose the
first architecture with a complexity linear to the number of nodes and edges
$O(N+E)$ by decoupling the local real-edge aggregation from the fully-connected
Transformer. We argue that this decoupling does not negatively affect the
expressivity, with our architecture being a universal function approximator for
graphs. Our GPS recipe consists of choosing 3 main ingredients: (i)
positional/structural encoding, (ii) local message-passing mechanism, and (iii)
global attention mechanism. We build and open-source a modular framework
$\textit{GraphGPS}$ that supports multiple types of encodings and that provides
efficiency and scalability both in small and large graphs. We test our
architecture on 11 benchmarks and show very competitive results on all of them,
show-casing the empirical benefits gained by the modularity and the combination
of different strategies. |
4,009 | MAVIPER: Learning Decision Tree Policies for Interpretable Multi-Agent Reinforcement Learning | Many recent breakthroughs in multi-agent reinforcement learning (MARL)
require the use of deep neural networks, which are challenging for human
experts to interpret and understand. On the other hand, existing work on
interpretable RL has shown promise in extracting more interpretable decision
tree-based policies, but only in the single-agent setting. To fill this gap, we
propose the first set of interpretable MARL algorithms that extract
decision-tree policies from neural networks trained with MARL. The first
algorithm, IVIPER, extends VIPER, a recent method for single-agent
interpretable RL, to the multi-agent setting. We demonstrate that IVIPER can
learn high-quality decision-tree policies for each agent. To better capture
coordination between agents, we propose a novel centralized decision-tree
training algorithm, MAVIPER. MAVIPER jointly grows the trees of each agent by
predicting the behavior of the other agents using their anticipated trees, and
uses resampling to focus on states that are critical for its interactions with
other agents. We show that both algorithms generally outperform the baselines
and that MAVIPER-trained agents achieve better-coordinated performance than
IVIPER-trained agents on three different multi-agent particle-world
environments. |
4,010 | Transportation-Inequalities, Lyapunov Stability and Sampling for Dynamical Systems on Continuous State Space | We study the concentration phenomenon for discrete-time random dynamical
systems with an unbounded state space. We develop a heuristic approach towards
obtaining exponential concentration inequalities for dynamical systems using an
entirely functional analytic framework. We also show that existence of
exponential-type Lyapunov function, compared to the purely deterministic
setting, not only implies stability but also exponential concentration
inequalities for sampling from the stationary distribution, via
\emph{transport-entropy inequality} (T-E). These results have significant
impact in \emph{reinforcement learning} (RL) and \emph{controls}, leading to
exponential concentration inequalities even for unbounded observables, while
neither assuming reversibility nor exact knowledge of random dynamical system
(assumptions at heart of concentration inequalities in statistical mechanics
and Markov diffusion processes). |
4,011 | FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech | We introduce FLEURS, the Few-shot Learning Evaluation of Universal
Representations of Speech benchmark. FLEURS is an n-way parallel speech dataset
in 102 languages built on top of the machine translation FLoRes-101 benchmark,
with approximately 12 hours of speech supervision per language. FLEURS can be
used for a variety of speech tasks, including Automatic Speech Recognition
(ASR), Speech Language Identification (Speech LangID), Translation and
Retrieval. In this paper, we provide baselines for the tasks based on
multilingual pre-trained models like mSLAM. The goal of FLEURS is to enable
speech technology in more languages and catalyze research in low-resource
speech understanding. |
4,012 | Over-the-Air Design of GAN Training for mmWave MIMO Channel Estimation | Future wireless systems are trending towards higher carrier frequencies that
offer larger communication bandwidth but necessitate the use of large antenna
arrays. Existing signal processing techniques for channel estimation do not
scale well to this "high-dimensional" regime in terms of performance and pilot
overhead. Meanwhile, training deep learning based approaches for channel
estimation requires large labeled datasets mapping pilot measurements to clean
channel realizations, which can only be generated offline using simulated
channels. In this paper, we develop a novel unsupervised over-the-air (OTA)
algorithm that utilizes noisy received pilot measurements to train a deep
generative model to output beamspace MIMO channel realizations. Our approach
leverages Generative Adversarial Networks (GAN), while using a conditional
input to distinguish between Line-of-Sight (LOS) and Non-Line-of-Sight (NLOS)
channel realizations. We also present a federated implementation of the OTA
algorithm that distributes the GAN training over multiple users and greatly
reduces the user side computation. We then formulate channel estimation from a
limited number of pilot measurements as an inverse problem and reconstruct the
channel by optimizing the input vector of the trained generative model. Our
proposed approach significantly outperforms Orthogonal Matching Pursuit on both
LOS and NLOS channel models, and EM-GM-AMP -- an Approximate Message Passing
algorithm -- on LOS channel models, while achieving comparable performance on
NLOS channel models in terms of the normalized channel reconstruction error.
More importantly, our proposed framework has the potential to be trained online
using real noisy pilot measurements, is not restricted to a specific channel
model and can even be utilized for a federated OTA design of a dataset
generator from noisy data. |
4,013 | Generating Natural Language Proofs with Verifier-Guided Search | Deductive reasoning (drawing conclusions from assumptions) is a challenging
problem in NLP. In this work, we focus on proof generation: given a hypothesis
and a set of supporting facts in natural language, the model generates a proof
tree indicating how to deduce the hypothesis from supporting facts. Instead of
generating the entire proof in one shot, prior work has demonstrated the
promise of stepwise generation but achieved limited success on real-world data.
Existing stepwise methods struggle to generate proof steps that are both valid
and relevant. In this paper, we present a novel stepwise method NLProofS
(Natural Language Proof Search), which learns to generate relevant steps
conditioning on the hypothesis. At the core of our approach, we train an
independent verifier to check the validity of proof steps. Instead of
generating steps greedily, we search for proofs maximizing a global proof score
judged by the verifier. NLProofS achieves state-of-the-art performance on
EntailmentBank and RuleTaker. For example, it improves the percentage of
correctly predicted proofs from 20.9% to 33.3% in the distractor setting of
EntailmentBank. This is the first time stepwise methods have led to better
generation of challenging human-authored proofs. |
4,014 | Lyapunov function approach for approximation algorithm design and analysis: with applications in submodular maximization | We propose a two-phase systematical framework for approximation algorithm
design and analysis via Lyapunov function. The first phase consists of using
Lyapunov function as an input and outputs a continuous-time approximation
algorithm with a provable approximation ratio. The second phase then converts
this continuous-time algorithm to a discrete-time algorithm with almost the
same approximation ratio along with provable time complexity. One distinctive
feature of our framework is that we only need to know the parametric form of
the Lyapunov function whose complete specification will not be decided until
the end of the first phase by maximizing the approximation ratio of the
continuous-time algorithm. Some immediate benefits of the Lyapunov function
approach include: (i) unifying many existing algorithms; (ii) providing a
guideline to design and analyze new algorithms; and (iii) offering new
perspectives to potentially improve existing algorithms. We use various
submodular maximization problems as running examples to illustrate our
framework. |
4,015 | Additive Logistic Mechanism for Privacy-Preserving Self-Supervised Learning | We study the privacy risks that are associated with training a neural
network's weights with self-supervised learning algorithms. Through empirical
evidence, we show that the fine-tuning stage, in which the network weights are
updated with an informative and often private dataset, is vulnerable to privacy
attacks. To address the vulnerabilities, we design a post-training
privacy-protection algorithm that adds noise to the fine-tuned weights and
propose a novel differential privacy mechanism that samples noise from the
logistic distribution. Compared to the two conventional additive noise
mechanisms, namely the Laplace and the Gaussian mechanisms, the proposed
mechanism uses a bell-shaped distribution that resembles the distribution of
the Gaussian mechanism, and it satisfies pure $\epsilon$-differential privacy
similar to the Laplace mechanism. We apply membership inference attacks on both
unprotected and protected models to quantify the trade-off between the models'
privacy and performance. We show that the proposed protection algorithm can
effectively reduce the attack accuracy to roughly 50\%-equivalent to random
guessing-while maintaining a performance loss below 5\%. |
4,016 | Towards Understanding Label Regularization for Fine-tuning Pre-trained Language Models | Knowledge Distillation (KD) is a prominent neural model compression technique
which heavily relies on teacher network predictions to guide the training of a
student model. Considering the ever-growing size of pre-trained language models
(PLMs), KD is often adopted in many NLP tasks involving PLMs. However, it is
evident that in KD, deploying the teacher network during training adds to the
memory and computational requirements of training. In the computer vision
literature, the necessity of the teacher network is put under scrutiny by
showing that KD is a label regularization technique that can be replaced with
lighter teacher-free variants such as the label-smoothing technique. However,
to the best of our knowledge, this issue is not investigated in NLP. Therefore,
this work concerns studying different label regularization techniques and
whether we actually need the teacher labels to fine-tune smaller PLM student
networks on downstream tasks. In this regard, we did a comprehensive set of
experiments on different PLMs such as BERT, RoBERTa, and GPT with more than 600
distinct trials and ran each configuration five times. This investigation led
to a surprising observation that KD and other label regularization techniques
do not play any meaningful role over regular fine-tuning when the student model
is pre-trained. We further explore this phenomenon in different settings of NLP
and computer vision tasks and demonstrate that pre-training itself acts as a
kind of regularization, and additional label regularization is unnecessary. |
4,017 | Non-stationary Bandits with Knapsacks | In this paper, we study the problem of bandits with knapsacks (BwK) in a
non-stationary environment. The BwK problem generalizes the multi-arm bandit
(MAB) problem to model the resource consumption associated with playing each
arm. At each time, the decision maker/player chooses to play an arm, and s/he
will receive a reward and consume certain amount of resource from each of the
multiple resource types. The objective is to maximize the cumulative reward
over a finite horizon subject to some knapsack constraints on the resources.
Existing works study the BwK problem under either a stochastic or adversarial
environment. Our paper considers a non-stationary environment which
continuously interpolates between these two extremes. We first show that the
traditional notion of variation budget is insufficient to characterize the
non-stationarity of the BwK problem for a sublinear regret due to the presence
of the constraints, and then we propose a new notion of global non-stationarity
measure. We employ both non-stationarity measures to derive upper and lower
bounds for the problem. Our results are based on a primal-dual analysis of the
underlying linear programs and highlight the interplay between the constraints
and the non-stationarity. Finally, we also extend the non-stationarity measure
to the problem of online convex optimization with constraints and obtain new
regret bounds accordingly. |
4,018 | Uniform Generalization Bound on Time and Inverse Temperature for Gradient Descent Algorithm and its Application to Analysis of Simulated Annealing | In this paper, we propose a novel uniform generalization bound on the time
and inverse temperature for stochastic gradient Langevin dynamics (SGLD) in a
non-convex setting. While previous works derive their generalization bounds by
uniform stability, we use Rademacher complexity to make our generalization
bound independent of the time and inverse temperature. Using Rademacher
complexity, we can reduce the problem to derive a generalization bound on the
whole space to that on a bounded region and therefore can remove the effect of
the time and inverse temperature from our generalization bound. As an
application of our generalization bound, an evaluation on the effectiveness of
the simulated annealing in a non-convex setting is also described. For the
sample size $n$ and time $s$, we derive evaluations with orders $\sqrt{n^{-1}
\log (n+1)}$ and $|(\log)^4(s)|^{-1}$, respectively. Here, $(\log)^4$ denotes
the $4$ times composition of the logarithmic function. |
4,019 | VulBERTa: Simplified Source Code Pre-Training for Vulnerability Detection | This paper presents VulBERTa, a deep learning approach to detect security
vulnerabilities in source code. Our approach pre-trains a RoBERTa model with a
custom tokenisation pipeline on real-world code from open-source C/C++
projects. The model learns a deep knowledge representation of the code syntax
and semantics, which we leverage to train vulnerability detection classifiers.
We evaluate our approach on binary and multi-class vulnerability detection
tasks across several datasets (Vuldeepecker, Draper, REVEAL and muVuldeepecker)
and benchmarks (CodeXGLUE and D2A). The evaluation results show that VulBERTa
achieves state-of-the-art performance and outperforms existing approaches
across different datasets, despite its conceptual simplicity, and limited cost
in terms of size of training data and number of model parameters. |
4,020 | Deletion and Insertion Tests in Regression Models | A basic task in explainable AI (XAI) is to identify the most important
features behind a prediction made by a black box function $f$. The insertion
and deletion tests of \cite{petsiuk2018rise} are used to judge the quality of
algorithms that rank pixels from most to least important for a classification.
Motivated by regression problems we establish a formula for their area under
the curve (AUC) criteria in terms of certain main effects and interactions in
an anchored decomposition of $f$. We find an expression for the expected value
of the AUC under a random ordering of inputs to $f$ and propose an alternative
area above a straight line for the regression setting. We use this criterion to
compare feature importances computed by integrated gradients (IG) to those
computed by Kernel SHAP (KS). Exact computation of KS grows exponentially with
dimension, while that of IG grows linearly with dimension. In two data sets
including binary variables we find that KS is superior to IG in insertion and
deletion tests, but only by a very small amount. Our comparison problems
include some binary inputs that pose a challenge to IG because it must use
values between the possible variable levels.
We show that IG will match KS when $f$ is an additive function plus a
multilinear function of the variables. This includes a multilinear
interpolation over the binary variables that would cause IG to have exponential
cost in a naive implementation. |
4,021 | Physics Guided Machine Learning for Variational Multiscale Reduced Order Modeling | We propose a new physics guided machine learning (PGML) paradigm that
leverages the variational multiscale (VMS) framework and available data to
dramatically increase the accuracy of reduced order models (ROMs) at a modest
computational cost. The hierarchical structure of the ROM basis and the VMS
framework enable a natural separation of the resolved and unresolved ROM
spatial scales. Modern PGML algorithms are used to construct novel models for
the interaction among the resolved and unresolved ROM scales. Specifically, the
new framework builds ROM operators that are closest to the true interaction
terms in the VMS framework. Finally, machine learning is used to reduce the
projection error and further increase the ROM accuracy. Our numerical
experiments for a two-dimensional vorticity transport problem show that the
novel PGML-VMS-ROM paradigm maintains the low computational cost of current
ROMs, while significantly increasing the ROM accuracy. |
4,022 | Tiered Reinforcement Learning: Pessimism in the Face of Uncertainty and Constant Regret | We propose a new learning framework that captures the tiered structure of
many real-world user-interaction applications, where the users can be divided
into two groups based on their different tolerance on exploration risks and
should be treated separately. In this setting, we simultaneously maintain two
policies $\pi^{\text{O}}$ and $\pi^{\text{E}}$: $\pi^{\text{O}}$ ("O" for
"online") interacts with more risk-tolerant users from the first tier and
minimizes regret by balancing exploration and exploitation as usual, while
$\pi^{\text{E}}$ ("E" for "exploit") exclusively focuses on exploitation for
risk-averse users from the second tier utilizing the data collected so far. An
important question is whether such a separation yields advantages over the
standard online setting (i.e., $\pi^{\text{E}}=\pi^{\text{O}}$) for the
risk-averse users. We individually consider the gap-independent
vs.~gap-dependent settings. For the former, we prove that the separation is
indeed not beneficial from a minimax perspective. For the latter, we show that
if choosing Pessimistic Value Iteration as the exploitation algorithm to
produce $\pi^{\text{E}}$, we can achieve a constant regret for risk-averse
users independent of the number of episodes $K$, which is in sharp contrast to
the $\Omega(\log K)$ regret for any online RL algorithms in the same setting,
while the regret of $\pi^{\text{O}}$ (almost) maintains its online regret
optimality and does not need to compromise for the success of $\pi^{\text{E}}$. |
4,023 | Differentially Private AUC Computation in Vertical Federated Learning | Federated learning has gained great attention recently as a privacy-enhancing
tool to jointly train a machine learning model by multiple parties. As a
sub-category, vertical federated learning (vFL) focuses on the scenario where
features and labels are split into different parties. The prior work on vFL has
mostly studied how to protect label privacy during model training. However,
model evaluation in vFL might also lead to potential leakage of private label
information. One mitigation strategy is to apply label differential privacy
(DP) but it gives bad estimations of the true (non-private) metrics. In this
work, we propose two evaluation algorithms that can more accurately compute the
widely used AUC (area under curve) metric when using label DP in vFL. Through
extensive experiments, we show our algorithms can achieve more accurate AUCs
compared to the baselines. |
4,024 | Linear Connectivity Reveals Generalization Strategies | It is widely accepted in the mode connectivity literature that when two
neural networks are trained similarly on the same data, they are connected by a
path through parameter space over which test set accuracy is maintained. Under
some circumstances, including transfer learning from pretrained models, these
paths are presumed to be linear. In contrast to existing results, we find that
among text classifiers (trained on MNLI, QQP, and CoLA), some pairs of
finetuned models have large barriers of increasing loss on the linear paths
between them. On each task, we find distinct clusters of models which are
linearly connected on the test loss surface, but are disconnected from models
outside the cluster -- models that occupy separate basins on the surface. By
measuring performance on specially-crafted diagnostic datasets, we find that
these clusters correspond to different generalization strategies: one cluster
behaves like a bag of words model under domain shift, while another cluster
uses syntactic heuristics. Our work demonstrates how the geometry of the loss
surface can guide models towards different heuristic functions. |
4,025 | AdaMix: Mixture-of-Adapter for Parameter-efficient Tuning of Large Language Models | Fine-tuning large-scale pre-trained language models to downstream tasks
require updating hundreds of millions of parameters. This not only increases
the serving cost to store a large copy of the model weights for every task, but
also exhibits instability during few-shot task adaptation. Parameter-efficient
techniques have been developed that tune small trainable components (e.g.,
adapters) injected in the large model while keeping most of the model weights
frozen. The prevalent mechanism to increase adapter capacity is to increase the
bottleneck dimension which increases the adapter parameters. In this work, we
introduce a new mechanism to improve adapter capacity without increasing
parameters or computational cost by two key techniques. (i) We introduce
multiple shared adapter components in each layer of the Transformer
architecture. We leverage sparse learning via random routing to update the
adapter parameters (encoder is kept frozen) resulting in the same amount of
computational cost (FLOPs) as that of training a single adapter. (ii) We
propose a simple merging mechanism to average the weights of multiple adapter
components to collapse to a single adapter in each Transformer layer, thereby,
keeping the overall parameters also the same but with significant performance
improvement. We demonstrate these techniques to work well across multiple task
settings including fully supervised and few-shot Natural Language Understanding
tasks. By only tuning 0.23% of a pre-trained language model's parameters, our
model outperforms the full model fine-tuning performance and several competing
methods. |
4,026 | Convolutional Neural Processes for Inpainting Satellite Images | The widespread availability of satellite images has allowed researchers to
model complex systems such as disease dynamics. However, many satellite images
have missing values due to measurement defects, which render them unusable
without data imputation. For example, the scanline corrector for the LANDSAT 7
satellite broke down in 2003, resulting in a loss of around 20\% of its data.
Inpainting involves predicting what is missing based on the known pixels and is
an old problem in image processing, classically based on PDEs or interpolation
methods, but recent deep learning approaches have shown promise. However, many
of these methods do not explicitly take into account the inherent
spatiotemporal structure of satellite images. In this work, we cast satellite
image inpainting as a natural meta-learning problem, and propose using
convolutional neural processes (ConvNPs) where we frame each satellite image as
its own task or 2D regression problem. We show ConvNPs can outperform classical
methods and state-of-the-art deep learning inpainting models on a scanline
inpainting problem for LANDSAT 7 satellite images, assessed on a variety of in
and out-of-distribution images. |
4,027 | Multi-Head Online Learning for Delayed Feedback Modeling | In online advertising, it is highly important to predict the probability and
the value of a conversion (e.g., a purchase). It not only impacts user
experience by showing relevant ads, but also affects ROI of advertisers and
revenue of marketplaces. Unlike clicks, which often occur within minutes after
impressions, conversions are expected to happen over a long period of time
(e.g., 30 days for online shopping). It creates a challenge, as the true labels
are only available after the long delays. Either inaccurate labels (partial
conversions) are used, or models are trained on stale data (e.g., from 30 days
ago). The problem is more eminent in online learning, which focuses on the live
performance on the latest data. In this paper, a novel solution is presented to
address this challenge using multi-head modeling. Unlike traditional methods,
it directly quantizes conversions into multiple windows, such as day 1, day 2,
day 3-7, and day 8-30. A sub-model is trained specifically on conversions
within each window. Label freshness is maximally preserved in early models
(e.g., day 1 and day 2), while late conversions are accurately utilized in
models with longer delays (e.g., day 8-30). It is shown to greatly exceed the
performance of known methods in online learning experiments for both conversion
rate (CVR) and value per click (VPC) predictions. Lastly, as a general method
for delayed feedback modeling, it can be combined with any advanced ML
techniques to further improve the performance. |
4,028 | Reward Uncertainty for Exploration in Preference-based Reinforcement Learning | Conveying complex objectives to reinforcement learning (RL) agents often
requires meticulous reward engineering. Preference-based RL methods are able to
learn a more flexible reward model based on human preferences by actively
incorporating human feedback, i.e. teacher's preferences between two clips of
behaviors. However, poor feedback-efficiency still remains a problem in current
preference-based RL algorithms, as tailored human feedback is very expensive.
To handle this issue, previous methods have mainly focused on improving query
selection and policy initialization. At the same time, recent exploration
methods have proven to be a recipe for improving sample-efficiency in RL. We
present an exploration method specifically for preference-based RL algorithms.
Our main idea is to design an intrinsic reward by measuring the novelty based
on learned reward. Specifically, we utilize disagreement across ensemble of
learned reward models. Our intuition is that disagreement in learned reward
model reflects uncertainty in tailored human feedback and could be useful for
exploration. Our experiments show that exploration bonus from uncertainty in
learned reward improves both feedback- and sample-efficiency of
preference-based RL algorithms on complex robot manipulation tasks from
MetaWorld benchmarks, compared with other existing exploration methods that
measure the novelty of state visitation. |
4,029 | Symbol Emergence as Inter-personal Categorization with Head-to-head Latent Word | In this study, we propose a head-to-head type (H2H-type) inter-personal
multimodal Dirichlet mixture (Inter-MDM) by modifying the original Inter-MDM,
which is a probabilistic generative model that represents the symbol emergence
between two agents as multiagent multimodal categorization. A
Metropolis--Hastings method-based naming game based on the Inter-MDM enables
two agents to collaboratively perform multimodal categorization and share signs
with a solid mathematical foundation of convergence. However, the conventional
Inter-MDM presumes a tail-to-tail connection across a latent word variable,
causing inflexibility of the further extension of Inter-MDM for modeling a more
complex symbol emergence. Therefore, we propose herein a head-to-head type
(H2H-type) Inter-MDM that treats a latent word variable as a child node of an
internal variable of each agent in the same way as many prior studies of
multimodal categorization. On the basis of the H2H-type Inter-MDM, we propose a
naming game in the same way as the conventional Inter-MDM. The experimental
results show that the H2H-type Inter-MDM yields almost the same performance as
the conventional Inter-MDM from the viewpoint of multimodal categorization and
sign sharing. |
4,030 | Sparse Mixers: Combining MoE and Mixing to build a more efficient BERT | We combine the capacity of sparsely gated Mixture-of-Experts (MoE) with the
speed and stability of linear, mixing transformations to design the Sparse
Mixer encoder model. The Sparse Mixer slightly outperforms (<1%) BERT on GLUE
and SuperGLUE, but more importantly trains 65% faster and runs inference 61%
faster. We also present a faster variant, prosaically named Fast Sparse Mixer,
that marginally underperforms (<0.2%) BERT on SuperGLUE, but trains and runs
nearly twice as fast: 89% faster training and 98% faster inference. We justify
the design of these two models by carefully ablating through various mixing
mechanisms, MoE configurations and model hyperparameters. The Sparse Mixer
overcomes many of the latency and stability concerns of MoE models and offers
the prospect of serving sparse student models, without resorting to distilling
them to dense variants. |
4,031 | Recipe2Vec: Multi-modal Recipe Representation Learning with Graph Neural Networks | Learning effective recipe representations is essential in food studies.
Unlike what has been developed for image-based recipe retrieval or learning
structural text embeddings, the combined effect of multi-modal information
(i.e., recipe images, text, and relation data) receives less attention. In this
paper, we formalize the problem of multi-modal recipe representation learning
to integrate the visual, textual, and relational information into recipe
embeddings. In particular, we first present Large-RG, a new recipe graph data
with over half a million nodes, making it the largest recipe graph to date. We
then propose Recipe2Vec, a novel graph neural network based recipe embedding
model to capture multi-modal information. Additionally, we introduce an
adversarial attack strategy to ensure stable learning and improve performance.
Finally, we design a joint objective function of node classification and
adversarial learning to optimize the model. Extensive experiments demonstrate
that Recipe2Vec outperforms state-of-the-art baselines on two classic food
study tasks, i.e., cuisine category classification and region prediction.
Dataset and codes are available at https://github.com/meettyj/Recipe2Vec. |
4,032 | PLAtE: A Large-scale Dataset for List Page Web Extraction | Recently, neural models have been leveraged to significantly improve the
performance of information extraction from semi-structured websites. However, a
barrier for continued progress is the small number of datasets large enough to
train these models. In this work, we introduce the PLAtE (Pages of Lists
Attribute Extraction) dataset as a challenging new web extraction task. PLAtE
focuses on shopping data, specifically extractions from product review pages
with multiple items. PLAtE encompasses both the tasks of: (1) finding
product-list segmentation boundaries and (2) extracting attributes for each
product. PLAtE is composed of 53, 905 items from 6, 810 pages, making it the
first large-scale list page web extraction dataset. We construct PLAtE by
collecting list pages from Common Crawl, then annotating them on Mechanical
Turk. Quantitative and qualitative analyses are performed to demonstrate PLAtE
has high-quality annotations. We establish strong baseline performance on PLAtE
with a SOTA model achieving an F1-score of 0.750 for attribute classification
and 0.915 for segmentation, indicating opportunities for future research
innovations in web extraction. |
4,033 | RecipeRec: A Heterogeneous Graph Learning Model for Recipe Recommendation | Recipe recommendation systems play an essential role in helping people decide
what to eat. Existing recipe recommendation systems typically focused on
content-based or collaborative filtering approaches, ignoring the higher-order
collaborative signal such as relational structure information among users,
recipes and food items. In this paper, we formalize the problem of recipe
recommendation with graphs to incorporate the collaborative signal into recipe
recommendation through graph modeling. In particular, we first present
URI-Graph, a new and large-scale user-recipe-ingredient graph. We then propose
RecipeRec, a novel heterogeneous graph learning model for recipe
recommendation. The proposed model can capture recipe content and collaborative
signal through a heterogeneous graph neural network with hierarchical attention
and an ingredient set transformer. We also introduce a graph contrastive
augmentation strategy to extract informative graph knowledge in a
self-supervised manner. Finally, we design a joint objective function of
recommendation and contrastive learning to optimize the model. Extensive
experiments demonstrate that RecipeRec outperforms state-of-the-art methods for
recipe recommendation. Dataset and codes are available at
https://github.com/meettyj/RecipeRec. |
4,034 | First Contact: Unsupervised Human-Machine Co-Adaptation via Mutual Information Maximization | How can we train an assistive human-machine interface (e.g., an
electromyography-based limb prosthesis) to translate a user's raw command
signals into the actions of a robot or computer when there is no prior mapping,
we cannot ask the user for supervision in the form of action labels or reward
feedback, and we do not have prior knowledge of the tasks the user is trying to
accomplish? The key idea in this paper is that, regardless of the task, when an
interface is more intuitive, the user's commands are less noisy. We formalize
this idea as a completely unsupervised objective for optimizing interfaces: the
mutual information between the user's command signals and the induced state
transitions in the environment. To evaluate whether this mutual information
score can distinguish between effective and ineffective interfaces, we conduct
an observational study on 540K examples of users operating various keyboard and
eye gaze interfaces for typing, controlling simulated robots, and playing video
games. The results show that our mutual information scores are predictive of
the ground-truth task completion metrics in a variety of domains, with an
average Spearman's rank correlation of 0.43. In addition to offline evaluation
of existing interfaces, we use our unsupervised objective to learn an interface
from scratch: we randomly initialize the interface, have the user attempt to
perform their desired tasks using the interface, measure the mutual information
score, and update the interface to maximize mutual information through
reinforcement learning. We evaluate our method through a user study with 12
participants who perform a 2D cursor control task using a perturbed mouse, and
an experiment with one user playing the Lunar Lander game using hand gestures.
The results show that we can learn an interface from scratch, without any user
supervision or prior knowledge of tasks, in under 30 minutes. |
4,035 | Imposing Gaussian Pre-Activations in a Neural Network | The goal of the present work is to propose a way to modify both the
initialization distribution of the weights of a neural network and its
activation function, such that all pre-activations are Gaussian. We propose a
family of pairs initialization/activation, where the activation functions span
a continuum from bounded functions (such as Heaviside or tanh) to the identity
function.
This work is motivated by the contradiction between existing works dealing
with Gaussian pre-activations: on one side, the works in the line of the Neural
Tangent Kernels and the Edge of Chaos are assuming it, while on the other side,
theoretical and experimental results challenge this hypothesis.
The family of pairs initialization/activation we are proposing will help us
to answer this hot question: is it desirable to have Gaussian pre-activations
in a neural network? |
4,036 | Hardness of Maximum Likelihood Learning of DPPs | Determinantal Point Processes (DPPs) are a widely used probabilistic model
for negatively correlated sets. DPPs have been successfully employed in Machine
Learning applications to select a diverse, yet representative subset of data.
In seminal work on DPPs in Machine Learning, Kulesza conjectured in his PhD
Thesis (2011) that the problem of finding a maximum likelihood DPP model for a
given data set is NP-complete.
In this work we prove Kulesza's conjecture. In fact, we prove the following
stronger hardness of approximation result: even computing a
$\left(1-O(\frac{1}{\log^9{N}})\right)$-approximation to the maximum
log-likelihood of a DPP on a ground set of $N$ elements is NP-complete. At the
same time, we also obtain the first polynomial-time algorithm that achieves a
nontrivial worst-case approximation to the optimal log-likelihood: the
approximation factor is $\frac{1}{(1+o(1))\log{m}}$ unconditionally (for data
sets that consist of $m$ subsets), and can be improved to $1-\frac{1+o(1)}{\log
N}$ if all $N$ elements appear in a $O(1/N)$-fraction of the subsets.
In terms of techniques, we reduce approximating the maximum log-likelihood of
DPPs on a data set to solving a gap instance of a "vector coloring" problem on
a hypergraph. Such a hypergraph is built on a bounded-degree graph construction
of Bogdanov, Obata and Trevisan (FOCS 2002), and is further enhanced by the
strong expanders of Alon and Capalbo (FOCS 2007) to serve our purposes. |
4,037 | Learning to Model Editing Processes | Most existing sequence generation models produce outputs in one pass, usually
left-to-right. However, this is in contrast with a more natural approach that
humans use in generating content; iterative refinement and editing. Recent work
has introduced edit-based models for various tasks (such as neural machine
translation and text style transfer), but these generally model a single edit
step. In this work, we propose modeling editing processes, modeling the whole
process of iteratively generating sequences. We form a conceptual framework to
describe the likelihood of multi-step edits, and describe neural models that
can learn a generative model of sequences based on these multistep edits. We
introduce baseline results and metrics on this task, finding that modeling
editing processes improves performance on a variety of axes on both our
proposed task and related downstream tasks compared to previous single-step
models of edits. |
4,038 | TorchNTK: A Library for Calculation of Neural Tangent Kernels of PyTorch Models | We introduce torchNTK, a python library to calculate the empirical neural
tangent kernel (NTK) of neural network models in the PyTorch framework. We
provide an efficient method to calculate the NTK of multilayer perceptrons. We
compare the explicit differentiation implementation against autodifferentiation
implementations, which have the benefit of extending the utility of the library
to any architecture supported by PyTorch, such as convolutional networks. A
feature of the library is that we expose the user to layerwise NTK components,
and show that in some regimes a layerwise calculation is more memory efficient.
We conduct preliminary experiments to demonstrate use cases for the software
and probe the NTK. |
4,039 | Low-rank Optimal Transport: Approximation, Statistics and Debiasing | The matching principles behind optimal transport (OT) play an increasingly
important role in machine learning, a trend which can be observed when OT is
used to disambiguate datasets in applications (e.g. single-cell genomics) or
used to improve more complex methods (e.g. balanced attention in transformers
or self-supervised learning). To scale to more challenging problems, there is a
growing consensus that OT requires solvers that can operate on millions, not
thousands, of points. The low-rank optimal transport (LOT) approach advocated
in \cite{scetbon2021lowrank} holds several promises in that regard, and was
shown to complement more established entropic regularization approaches, being
able to insert itself in more complex pipelines, such as quadratic OT. LOT
restricts the search for low-cost couplings to those that have a
low-nonnegative rank, yielding linear time algorithms in cases of interest.
However, these promises can only be fulfilled if the LOT approach is seen as a
legitimate contender to entropic regularization when compared on properties of
interest, where the scorecard typically includes theoretical properties
(statistical bounds, relation to other methods) or practical aspects
(debiasing, hyperparameter tuning, initialization). We target each of these
areas in this paper in order to cement the impact of low-rank approaches in
computational OT. |
4,040 | Wavelet Feature Maps Compression for Image-to-Image CNNs | Convolutional Neural Networks (CNNs) are known for requiring extensive
computational resources, and quantization is among the best and most common
methods for compressing them. While aggressive quantization (i.e., less than
4-bits) performs well for classification, it may cause severe performance
degradation in image-to-image tasks such as semantic segmentation and depth
estimation. In this paper, we propose Wavelet Compressed Convolution (WCC) -- a
novel approach for high-resolution activation maps compression integrated with
point-wise convolutions, which are the main computational cost of modern
architectures. To this end, we use an efficient and hardware-friendly
Haar-wavelet transform, known for its effectiveness in image compression, and
define the convolution on the compressed activation map. We experiment on
various tasks, that benefit from high-resolution input, and by combining WCC
with light quantization, we achieve compression rates equivalent to 1-4bit
activation quantization with relatively small and much more graceful
degradation in performance. |
4,041 | Accelerating hydrodynamic simulations of urban drainage systems with physics-guided machine learning | We propose and demonstrate a new approach for fast and accurate surrogate
modelling of urban drainage system hydraulics based on physics-guided machine
learning. The surrogates are trained against a limited set of simulation
results from a hydrodynamic (HiFi) model. Our approach reduces simulation times
by one to two orders of magnitude compared to a HiFi model. It is thus slower
than e.g. conceptual hydrological models, but it enables simulations of water
levels, flows and surcharges in all nodes and links of a drainage network and
thus largely preserves the level of detail provided by HiFi models. Comparing
time series simulated by the surrogate and the HiFi model, R2 values in the
order of 0.9 are achieved. Surrogate training times are currently in the order
of one hour. However, they can likely be reduced through the application of
transfer learning and graph neural networks. Our surrogate approach will be
useful for interactive workshops in initial design phases of urban drainage
systems, as well as for real time applications. In addition, our model
formulation is generic and future research should investigate its application
for simulating other water systems. |
4,042 | K-12BERT: BERT for K-12 education | Online education platforms are powered by various NLP pipelines, which
utilize models like BERT to aid in content curation. Since the inception of the
pre-trained language models like BERT, there have also been many efforts toward
adapting these pre-trained models to specific domains. However, there has not
been a model specifically adapted for the education domain (particularly K-12)
across subjects to the best of our knowledge. In this work, we propose to train
a language model on a corpus of data curated by us across multiple subjects
from various sources for K-12 education. We also evaluate our model, K12-BERT,
on downstream tasks like hierarchical taxonomy tagging. |
4,043 | Certified Robustness Against Natural Language Attacks by Causal Intervention | Deep learning models have achieved great success in many fields, yet they are
vulnerable to adversarial examples. This paper follows a causal perspective to
look into the adversarial vulnerability and proposes Causal Intervention by
Semantic Smoothing (CISS), a novel framework towards robustness against natural
language attacks. Instead of merely fitting observational data, CISS learns
causal effects p(y|do(x)) by smoothing in the latent semantic space to make
robust predictions, which scales to deep architectures and avoids tedious
construction of noise customized for specific attacks. CISS is provably robust
against word substitution attacks, as well as empirically robust even when
perturbations are strengthened by unknown attack algorithms. For example, on
YELP, CISS surpasses the runner-up by 6.7% in terms of certified robustness
against word substitutions, and achieves 79.4% empirical robustness when
syntactic attacks are integrated. |
4,044 | Beyond Impossibility: Balancing Sufficiency, Separation and Accuracy | Among the various aspects of algorithmic fairness studied in recent years,
the tension between satisfying both \textit{sufficiency} and
\textit{separation} -- e.g. the ratios of positive or negative predictive
values, and false positive or false negative rates across groups -- has
received much attention. Following a debate sparked by COMPAS, a criminal
justice predictive system, the academic community has responded by laying out
important theoretical understanding, showing that one cannot achieve both with
an imperfect predictor when there is no equal distribution of labels across the
groups. In this paper, we shed more light on what might be still possible
beyond the impossibility -- the existence of a trade-off means we should aim to
find a good balance within it. After refining the existing theoretical result,
we propose an objective that aims to balance \textit{sufficiency} and
\textit{separation} measures, while maintaining similar accuracy levels. We
show the use of such an objective in two empirical case studies, one involving
a multi-objective framework, and the other fine-tuning of a model pre-trained
for accuracy. We show promising results, where better trade-offs are achieved
compared to existing alternatives. |
4,045 | ColdGuess: A General and Effective Relational Graph Convolutional Network to Tackle Cold Start Cases | Low-quality listings and bad actor behavior in online retail websites
threatens e-commerce business as these result in sub-optimal buying experience
and erode customer trust. When a new listing is created, how to tell it has
good-quality? Is the method effective, fast, and scalable? Previous approaches
often have three limitations/challenges: (1) unable to handle cold start
problems where new sellers/listings lack sufficient selling histories. (2)
inability of scoring hundreds of millions of listings at scale, or compromise
performance for scalability. (3) has space challenges from large-scale graph
with giant e-commerce business size. To overcome these limitations/challenges,
we proposed ColdGuess, an inductive graph-based risk predictor built upon a
heterogeneous seller product graph, which effectively identifies risky
seller/product/listings at scale. ColdGuess tackles the large-scale graph by
consolidated nodes, and addresses the cold start problems using homogeneous
influence1. The evaluation on real data demonstrates that ColdGuess has stable
performance as the number of unknown features increases. It outperforms the
lightgbm2 by up to 34 pcp ROC-AUC in a cold start case when a new seller sells
a new product . The resulting system, ColdGuess, is effective, adaptable to
changing risky seller behavior, and is already in production |
4,046 | Fast & Furious: Modelling Malware Detection as Evolving Data Streams | Malware is a major threat to computer systems and imposes many challenges to
cyber security. Targeted threats, such as ransomware, cause millions of dollars
in losses every year. The constant increase of malware infections has been
motivating popular antiviruses (AVs) to develop dedicated detection strategies,
which include meticulously crafted machine learning (ML) pipelines. However,
malware developers unceasingly change their samples features to bypass
detection. This constant evolution of malware samples causes changes to the
data distribution (i.e., concept drifts) that directly affect ML model
detection rates. In this work, we evaluate the impact of concept drift on
malware classifiers for two Android datasets: DREBIN (~130K apps) and AndroZoo
(~350K apps). Android is a ubiquitous operating system for smartphones, which
stimulates attackers to regularly create and update malware to the platform. We
conducted a longitudinal evaluation by (i) classifying malware samples
collected over nine years (2009-2018), (ii) reviewing concept drift detection
algorithms to attest its pervasiveness, (iii) comparing distinct ML approaches
to mitigate the issue, and (iv) proposing an ML data stream pipeline that
outperformed literature approaches. As a result, we observed that updating
every component of the pipeline in response to concept drifts allows the
classification model to achieve increasing detection rates as the data
representation (extracted features) is updated. Furthermore, we discuss the
impact of the changes on the classification models by comparing the variations
in the extracted features. |
4,047 | FreDo: Frequency Domain-based Long-Term Time Series Forecasting | The ability to forecast far into the future is highly beneficial to many
applications, including but not limited to climatology, energy consumption, and
logistics. However, due to noise or measurement error, it is questionable how
far into the future one can reasonably predict. In this paper, we first
mathematically show that due to error accumulation, sophisticated models might
not outperform baseline models for long-term forecasting. To demonstrate, we
show that a non-parametric baseline model based on periodicity can actually
achieve comparable performance to a state-of-the-art Transformer-based model on
various datasets. We further propose FreDo, a frequency domain-based neural
network model that is built on top of the baseline model to enhance its
performance and which greatly outperforms the state-of-the-art model. Finally,
we validate that the frequency domain is indeed better by comparing univariate
models trained in the frequency v.s. time domain. |
4,048 | lpSpikeCon: Enabling Low-Precision Spiking Neural Network Processing for Efficient Unsupervised Continual Learning on Autonomous Agents | Recent advances have shown that SNN-based systems can efficiently perform
unsupervised continual learning due to their bio-plausible learning rule, e.g.,
Spike-Timing-Dependent Plasticity (STDP). Such learning capabilities are
especially beneficial for use cases like autonomous agents (e.g., robots and
UAVs) that need to continuously adapt to dynamically changing
scenarios/environments, where new data gathered directly from the environment
may have novel features that should be learned online. Current state-of-the-art
works employ high-precision weights (i.e., 32 bit) for both training and
inference phases, which pose high memory and energy costs thereby hindering
efficient embedded implementations of such systems for battery-driven mobile
autonomous systems. On the other hand, precision reduction may jeopardize the
quality of unsupervised continual learning due to information loss. Towards
this, we propose lpSpikeCon, a novel methodology to enable low-precision SNN
processing for efficient unsupervised continual learning on
resource-constrained autonomous agents/systems. Our lpSpikeCon methodology
employs the following key steps: (1) analyzing the impacts of training the SNN
model under unsupervised continual learning settings with reduced weight
precision on the inference accuracy; (2) leveraging this study to identify SNN
parameters that have a significant impact on the inference accuracy; and (3)
developing an algorithm for searching the respective SNN parameter values that
improve the quality of unsupervised continual learning. The experimental
results show that our lpSpikeCon can reduce weight memory of the SNN model by
8x (i.e., by judiciously employing 4-bit weights) for performing online
training with unsupervised continual learning and achieve no accuracy loss in
the inference phase, as compared to the baseline model with 32-bit weights
across different network sizes. |
4,049 | Policy Compliance Detection via Expression Tree Inference | Policy Compliance Detection (PCD) is a task we encounter when reasoning over
texts, e.g. legal frameworks. Previous work to address PCD relies heavily on
modeling the task as a special case of Recognizing Textual Entailment.
Entailment is applicable to the problem of PCD, however viewing the policy as a
single proposition, as opposed to multiple interlinked propositions, yields
poor performance and lacks explainability. To address this challenge, more
recent proposals for PCD have argued for decomposing policies into expression
trees consisting of questions connected with logic operators. Question
answering is used to obtain answers to these questions with respect to a
scenario. Finally, the expression tree is evaluated in order to arrive at an
overall solution. However, this work assumes expression trees are provided by
experts, thus limiting its applicability to new policies. In this work, we
learn how to infer expression trees automatically from policy texts. We ensure
the validity of the inferred trees by introducing constrained decoding using a
finite state automaton to ensure the generation of valid trees. We determine
through automatic evaluation that 63% of the expression trees generated by our
constrained generation model are logically equivalent to gold trees. Human
evaluation shows that 88% of trees generated by our model are correct. |
4,050 | History Compression via Language Models in Reinforcement Learning | In a partially observable Markov decision process (POMDP), an agent typically
uses a representation of the past to approximate the underlying MDP. We propose
to utilize a frozen Pretrained Language Transformer (PLT) for history
representation and compression to improve sample efficiency. To avoid training
of the Transformer, we introduce FrozenHopfield, which automatically associates
observations with pretrained token embeddings. To form these associations, a
modern Hopfield network stores these token embeddings, which are retrieved by
queries that are obtained by a random but fixed projection of observations. Our
new method, HELM, enables actor-critic network architectures that contain a
pretrained language Transformer for history representation as a memory module.
Since a representation of the past need not be learned, HELM is much more
sample efficient than competitors. On Minigrid and Procgen environments HELM
achieves new state-of-the-art results. Our code is available at
https://github.com/ml-jku/helm. |
4,051 | Interpretation Quality Score for Measuring the Quality of interpretability methods | Machine learning (ML) models have been applied to a wide range of natural
language processing (NLP) tasks in recent years. In addition to making accurate
decisions, the necessity of understanding how models make their decisions has
become apparent in many applications. To that end, many interpretability
methods that help explain the decision processes of ML models have been
developed. Yet, there currently exists no widely-accepted metric to evaluate
the quality of explanations generated by these methods. As a result, there
currently is no standard way of measuring to what degree an interpretability
method achieves an intended objective. Moreover, there is no accepted standard
of performance by which we can compare and rank the current existing
interpretability methods. In this paper, we propose a novel metric for
quantifying the quality of explanations generated by interpretability methods.
We compute the metric on three NLP tasks using six interpretability methods and
present our results. |
4,052 | Taming the sign problem of explicitly antisymmetrized neural networks via rough activation functions | Explicit antisymmetrization of a two-layer neural network is a potential
candidate for a universal function approximator for generic antisymmetric
functions, which are ubiquitous in quantum physics. However, this strategy
suffers from a sign problem, namely, due to near exact cancellation of positive
and negative contributions, the magnitude of the antisymmetrized function may
be significantly smaller than that before antisymmetrization. We prove that the
severity of the sign problem is directly related to the smoothness of the
activation function. For smooth activation functions (e.g., $\tanh$), the sign
problem of the explicitly antisymmetrized two-layer neural network deteriorates
super-polynomially with respect to the system size. On the other hand, for
rough activation functions (e.g., ReLU), the deterioration rate of the sign
problem can be tamed to be at most polynomial with respect to the system size.
Finally, the cost of a direct implementation of antisymmetrized two-layer
neural network scales factorially with respect to the system size. We describe
an efficient algorithm for approximate evaluation of such a network, of which
the cost scales polynomially with respect to the system size and inverse
precision. |
4,053 | RevUp: Revise and Update Information Bottleneck for Event Representation | In machine learning, latent variables play a key role to capture the
underlying structure of data, but they are often unsupervised. When we have
side knowledge that already has high-level information about the input data, we
can use that source to guide latent variables and capture the available
background information in a process called "parameter injection." In that
regard, we propose a semi-supervised information bottleneck-based model that
enables the use of side knowledge, even if it is noisy and imperfect, to direct
the learning of discrete latent variables. Fundamentally, we introduce an
auxiliary continuous latent variable as a way to reparameterize the model's
discrete variables with a light-weight hierarchical structure. With this
reparameterization, the model's discrete latent variables are learned to
minimize the mutual information between the observed data and optional side
knowledge that is not already captured by the new, auxiliary variables. We
theoretically show that our approach generalizes an existing method of
parameter injection, and perform an empirical case study of our approach on
language-based event modeling. We corroborate our theoretical results with
strong empirical experiments, showing that the proposed method outperforms
previous proposed approaches on multiple datasets. |
4,054 | Asynchronous Neural Networks for Learning in Graphs | This paper studies asynchronous message passing (AMP), a new paradigm for
applying neural network based learning to graphs. Existing graph neural
networks use the synchronous distributed computing model and aggregate their
neighbors in each round, which causes problems such as oversmoothing and limits
their expressiveness. On the other hand, AMP is based on the asynchronous
model, where nodes react to messages of their neighbors individually. We prove
that (i) AMP can simulate synchronous GNNs and that (ii) AMP can theoretically
distinguish any pair of graphs. We experimentally validate AMP's
expressiveness. Further, we show that AMP might be better suited to propagate
messages over large distances in graphs and performs well on several graph
classification benchmarks. |
4,055 | EBM Life Cycle: MCMC Strategies for Synthesis, Defense, and Density Modeling | This work presents strategies to learn an Energy-Based Model (EBM) according
to the desired length of its MCMC sampling trajectories. MCMC trajectories of
different lengths correspond to models with different purposes. Our experiments
cover three different trajectory magnitudes and learning outcomes: 1) shortrun
sampling for image generation; 2) midrun sampling for classifier-agnostic
adversarial defense; and 3) longrun sampling for principled modeling of image
probability densities. To achieve these outcomes, we introduce three novel
methods of MCMC initialization for negative samples used in Maximum Likelihood
(ML) learning. With standard network architectures and an unaltered ML
objective, our MCMC initialization methods alone enable significant performance
gains across the three applications that we investigate. Our results include
state-of-the-art FID scores for unnormalized image densities on the CIFAR-10
and ImageNet datasets; state-of-the-art adversarial defense on CIFAR-10 among
purification methods and the first EBM defense on ImageNet; and scalable
techniques for learning valid probability densities. Code for this project can
be found at https://github.com/point0bar1/ebm-life-cycle. |
4,056 | Gacs-Korner Common Information Variational Autoencoder | We propose a notion of common information that allows one to quantify and
separate the information that is shared between two random variables from the
information that is unique to each. Our notion of common information is a
variational relaxation of the G\'acs-K\"orner common information, which we
recover as a special case, but is more amenable to optimization and can be
approximated empirically using samples from the underlying distribution. We
then provide a method to partition and quantify the common and unique
information using a simple modification of a traditional variational
auto-encoder. Empirically, we demonstrate that our formulation allows us to
learn semantically meaningful common and unique factors of variation even on
high-dimensional data such as images and videos. Moreover, on datasets where
ground-truth latent factors are known, we show that we can accurately quantify
the common information between the random variables. Additionally, we show that
the auto-encoder that we learn recovers semantically meaningful disentangled
factors of variation, even though we do not explicitly optimize for it. |
4,057 | Psychotic Relapse Prediction in Schizophrenia Patients using A Mobile Sensing-based Supervised Deep Learning Model | Mobile sensing-based modeling of behavioral changes could predict an oncoming
psychotic relapse in schizophrenia patients for timely interventions. Deep
learning models could complement existing non-deep learning models for relapse
prediction by modeling latent behavioral features relevant to the prediction.
However, given the inter-individual behavioral differences, model
personalization might be required for a predictive model. In this work, we
propose RelapsePredNet, a Long Short-Term Memory (LSTM) neural network-based
model for relapse prediction. The model is personalized for a particular
patient by training using data from patients most similar to the given patient.
Several demographics and baseline mental health scores were considered as
personalization metrics to define patient similarity. We investigated the
effect of personalization on training dataset characteristics, learned
embeddings, and relapse prediction performance. We compared RelapsePredNet with
a deep learning-based anomaly detection model for relapse prediction. Further,
we investigated if RelapsePredNet could complement ClusterRFModel (a random
forest model leveraging clustering and template features proposed in prior
work) in a fusion model, by identifying latent behavioral features relevant for
relapse prediction. The CrossCheck dataset consisting of continuous mobile
sensing data obtained from 63 schizophrenia patients, each monitored for up to
a year, was used for our evaluations. The proposed RelapsePredNet outperformed
the deep learning-based anomaly detection model for relapse prediction. The F2
score for prediction were 0.21 and 0.52 in the full test set and the Relapse
Test Set (consisting of data from patients who have had relapse only),
respectively. These corresponded to a 29.4% and 38.8% improvement compared to
the existing deep learning-based model for relapse prediction. |
4,058 | Forecasting Multilinear Data via Transform-Based Tensor Autoregression | In the era of big data, there is an increasing demand for new methods for
analyzing and forecasting 2-dimensional data. The current research aims to
accomplish these goals through the combination of time-series modeling and
multilinear algebraic systems. We expand previous autoregressive techniques to
forecast multilinear data, aptly named the L-Transform Tensor autoregressive
(L-TAR for short). Tensor decompositions and multilinear tensor products have
allowed for this approach to be a feasible method of forecasting. We achieve
statistical independence between the columns of the observations through
invertible discrete linear transforms, enabling a divide and conquer approach.
We present an experimental validation of the proposed methods on datasets
containing image collections, video sequences, sea surface temperature
measurements, stock prices, and networks. |
4,059 | Rethinking Evaluation Practices in Visual Question Answering: A Case Study on Out-of-Distribution Generalization | Vision-and-language (V&L) models pretrained on large-scale multimodal data
have demonstrated strong performance on various tasks such as image captioning
and visual question answering (VQA). The quality of such models is commonly
assessed by measuring their performance on unseen data that typically comes
from the same distribution as the training data. However, we observe that these
models exhibit poor out-of-distribution (OOD) generalization on the task of
VQA. To better understand the underlying causes of poor generalization, we
comprehensively investigate performance of two pretrained V&L models under
different settings (i.e. classification and open-ended text generation) by
conducting cross-dataset evaluations. We find that these models tend to learn
to solve the benchmark, rather than learning the high-level skills required by
the VQA task. We also argue that in most cases generative models are less
susceptible to shifts in data distribution, while frequently performing better
on our tested benchmarks. Moreover, we find that multimodal pretraining
improves OOD performance in most settings. Finally, we revisit assumptions
underlying the use of automatic VQA evaluation metrics, and empirically show
that their stringent nature repeatedly penalizes models for correct responses. |
4,060 | Learning for Expressive Task-Related Sentence Representations | NLP models learn sentence representations for downstream tasks by tuning a
model which is pre-trained by masked language modeling. However, after tuning,
the learned sentence representations may be skewed heavily toward label space
and thus are not expressive enough to represent whole samples, which should
contain task-related information of both sentence inputs and labels. In this
work, we learn expressive sentence representations for supervised tasks which
(1). contain task-related information in the sentence inputs, and (2). enable
correct label predictions. To achieve this goal, we first propose a new
objective which explicitly points out the label token space in the input, and
predicts categories of labels via an added [MASK] token. This objective
encourages fusing the semantic information of both the label and sentence. Then
we develop a neighbor attention module, added on a frozen pre-trained model, to
build connections between label/sentence tokens via their neighbors. The
propagation can be further guided by the regularization on neighborhood
representations to encourage expressiveness. Experimental results show that,
despite tuning only 5% additional parameters over a frozen pre-trained model,
our model can achieve classification results comparable to the SOTA while
maintaining strong expressiveness as well. |
4,061 | Distributional Hamilton-Jacobi-Bellman Equations for Continuous-Time Reinforcement Learning | Continuous-time reinforcement learning offers an appealing formalism for
describing control problems in which the passage of time is not naturally
divided into discrete increments. Here we consider the problem of predicting
the distribution of returns obtained by an agent interacting in a
continuous-time, stochastic environment. Accurate return predictions have
proven useful for determining optimal policies for risk-sensitive control,
learning state representations, multiagent coordination, and more. We begin by
establishing the distributional analogue of the Hamilton-Jacobi-Bellman (HJB)
equation for It\^o diffusions and the broader class of Feller-Dynkin processes.
We then specialize this equation to the setting in which the return
distribution is approximated by $N$ uniformly-weighted particles, a common
design choice in distributional algorithms. Our derivation highlights
additional terms due to statistical diffusivity which arise from the proper
handling of distributions in the continuous-time setting. Based on this, we
propose a tractable algorithm for approximately solving the distributional HJB
based on a JKO scheme, which can be implemented in an online control algorithm.
We demonstrate the effectiveness of such an algorithm in a synthetic control
problem. |
4,062 | Regret-Aware Black-Box Optimization with Natural Gradients, Trust-Regions and Entropy Control | Most successful stochastic black-box optimizers, such as CMA-ES, use rankings
of the individual samples to obtain a new search distribution. Yet, the use of
rankings also introduces several issues such as the underlying optimization
objective is often unclear, i.e., we do not optimize the expected fitness.
Further, while these algorithms typically produce a high-quality mean estimate
of the search distribution, the produced samples can have poor quality as these
algorithms are ignorant of the regret. Lastly, noisy fitness function
evaluations may result in solutions that are highly sub-optimal on expectation.
In contrast, stochastic optimizers that are motivated by policy gradients, such
as the Model-based Relative Entropy Stochastic Search (MORE) algorithm,
directly optimize the expected fitness function without the use of rankings.
MORE can be derived by applying natural policy gradients and compatible
function approximation, and is using information theoretic constraints to
ensure the stability of the policy update. While MORE does not suffer from the
listed limitations, it often cannot achieve state of the art performance in
comparison to ranking based methods. We improve MORE by decoupling the update
of the mean and covariance of the search distribution allowing for more
aggressive updates on the mean while keeping the update on the covariance
conservative, an improved entropy scheduling technique based on an evolution
path which results in faster convergence and a simplified and more effective
model learning approach in comparison to the original paper. We compare our
algorithm to state of the art black-box optimization algorithms on standard
optimization tasks as well as on episodic RL tasks in robotics where it is also
crucial to have small regret. We obtain competitive results on benchmark
functions and clearly outperform ranking-based methods in terms of regret on
the RL tasks. |
4,063 | Byzantine Machine Learning Made Easy by Resilient Averaging of Momentums | Byzantine resilience emerged as a prominent topic within the distributed
machine learning community. Essentially, the goal is to enhance distributed
optimization algorithms, such as distributed SGD, in a way that guarantees
convergence despite the presence of some misbehaving (a.k.a., {\em Byzantine})
workers. Although a myriad of techniques addressing the problem have been
proposed, the field arguably rests on fragile foundations. These techniques are
hard to prove correct and rely on assumptions that are (a) quite unrealistic,
i.e., often violated in practice, and (b) heterogeneous, i.e., making it
difficult to compare approaches.
We present \emph{RESAM (RESilient Averaging of Momentums)}, a unified
framework that makes it simple to establish optimal Byzantine resilience,
relying only on standard machine learning assumptions. Our framework is mainly
composed of two operators: \emph{resilient averaging} at the server and
\emph{distributed momentum} at the workers. We prove a general theorem stating
the convergence of distributed SGD under RESAM. Interestingly, demonstrating
and comparing the convergence of many existing techniques become direct
corollaries of our theorem, without resorting to stringent assumptions. We also
present an empirical evaluation of the practical relevance of RESAM. |
4,064 | D$^\text{2}$UF: Deep Coded Aperture Design and Unrolling Algorithm for Compressive Spectral Image Fusion | Compressive spectral imaging (CSI) has attracted significant attention since
it employs synthetic apertures to codify spatial and spectral information,
sensing only 2D projections of the 3D spectral image. However, these optical
architectures suffer from a trade-off between the spatial and spectral
resolution of the reconstructed image due to technology limitations. To
overcome this issue, compressive spectral image fusion (CSIF) employs the
projected measurements of two CSI architectures with different resolutions to
estimate a high-spatial high-spectral resolution. This work presents the fusion
of the compressive measurements of a low-spatial high-spectral resolution coded
aperture snapshot spectral imager (CASSI) architecture and a high-spatial
low-spectral resolution multispectral color filter array (MCFA) system. Unlike
previous CSIF works, this paper proposes joint optimization of the sensing
architectures and a reconstruction network in an end-to-end (E2E) manner. The
trainable optical parameters are the coded aperture (CA) in the CASSI and the
colored coded aperture in the MCFA system, employing a sigmoid activation
function and regularization function to encourage binary values on the
trainable variables for an implementation purpose. Additionally, an
unrolling-based network inspired by the alternating direction method of
multipliers (ADMM) optimization is formulated to address the reconstruction
step and the acquisition systems design jointly. Finally, a spatial-spectral
inspired loss function is employed at the end of each unrolling layer to
increase the convergence of the unrolling network. The proposed method
outperforms previous CSIF methods, and experimental results validate the method
with real measurements. |
4,065 | Not too little, not too much: a theoretical analysis of graph (over)smoothing | We analyze graph smoothing with \emph{mean aggregation}, where each node
successively receives the average of the features of its neighbors. Indeed, it
has quickly been observed that Graph Neural Networks (GNNs), which generally
follow some variant of Message-Passing (MP) with repeated aggregation, may be
subject to the \emph{oversmoothing} phenomenon: by performing too many rounds
of MP, the node features tend to converge to a non-informative limit. In the
case of mean aggregation, for connected graphs, the node features become
constant across the whole graph. At the other end of the spectrum, it is
intuitively obvious that \emph{some} MP rounds are necessary, but existing
analyses do not exhibit both phenomena at once: beneficial ``finite'' smoothing
and oversmoothing in the limit. In this paper, we consider simplified linear
GNNs, and rigorously analyze two examples for which a finite number of mean
aggregation steps provably improves the learning performance, before
oversmoothing kicks in. We consider a latent space random graph model, where
node features are partial observations of the latent variables and the graph
contains pairwise relationships between them. We show that graph smoothing
restores some of the lost information, up to a certain point, by two
phenomenon: graph smoothing shrinks non-principal directions in the data faster
than principal ones, which is useful for regression, and shrinks nodes within
communities faster than they collapse together, which improves classification. |
4,066 | Mathematical Models of Human Drivers Using Artificial Risk Fields | In this paper, we use the concept of artificial risk fields to predict how
human operators control a vehicle in response to upcoming road situations. A
risk field assigns a non-negative risk measure to the state of the system in
order to model how close that state is to violating a safety property, such as
hitting an obstacle or exiting the road. Using risk fields, we construct a
stochastic model of the operator that maps from states to likely actions. We
demonstrate our approach on a driving task wherein human subjects are asked to
drive a car inside a realistic driving simulator while avoiding obstacles
placed on the road. We show that the most likely risk field given the driving
data is obtained by solving a convex optimization problem. Next, we apply the
inferred risk fields to generate distinct driving behaviors while comparing
predicted trajectories against ground truth measurements. We observe that the
risk fields are excellent at predicting future trajectory distributions with
high prediction accuracy for up to twenty seconds prediction horizons. At the
same time, we observe some challenges such as the inability to account for how
drivers choose to accelerate/decelerate based on the road conditions. |
4,067 | One-Pixel Shortcut: on the Learning Preference of Deep Neural Networks | Unlearnable examples (ULEs) aim to protect data from unauthorized usage for
training DNNs. Error-minimizing noise, which is injected to clean data, is one
of the most successful methods for preventing DNNs from giving correct
predictions on incoming new data. Nonetheless, under specific training
strategies such as adversarial training, the unlearnability of error-minimizing
noise will severely degrade. In addition, the transferability of
error-minimizing noise is inherently limited by the mismatch between the
generator model and the targeted learner model. In this paper, we investigate
the mechanism of unlearnable examples and propose a novel model-free method,
named \emph{One-Pixel Shortcut}, which only perturbs a single pixel of each
image and makes the dataset unlearnable. Our method needs much less
computational cost and obtains stronger transferability and thus can protect
data from a wide range of different models. Based on this, we further introduce
the first unlearnable dataset called CIFAR-10-S, which is indistinguishable
from normal CIFAR-10 by human observers and can serve as a benchmark for
different models or training strategies to evaluate their abilities to extract
critical features from the disturbance of non-semantic representations. The
original error-minimizing ULEs will lose efficiency under adversarial training,
where the model can get over 83\% clean test accuracy. Meanwhile, even if
adversarial training and strong data augmentation like RandAugment are applied
together, the model trained on CIFAR-10-S cannot get over 50\% clean test
accuracy. |
4,068 | Adversarial Attack on Attackers: Post-Process to Mitigate Black-Box Score-Based Query Attacks | The score-based query attacks (SQAs) pose practical threats to deep neural
networks by crafting adversarial perturbations within dozens of queries, only
using the model's output scores. Nonetheless, we note that if the loss trend of
the outputs is slightly perturbed, SQAs could be easily misled and thereby
become much less effective. Following this idea, we propose a novel defense,
namely Adversarial Attack on Attackers (AAA), to confound SQAs towards
incorrect attack directions by slightly modifying the output logits. In this
way, (1) SQAs are prevented regardless of the model's worst-case robustness;
(2) the original model predictions are hardly changed, i.e., no degradation on
clean accuracy; (3) the calibration of confidence scores can be improved
simultaneously. Extensive experiments are provided to verify the above
advantages. For example, by setting $\ell_\infty=8/255$ on CIFAR-10, our
proposed AAA helps WideResNet-28 secure $80.59\%$ accuracy under Square attack
($2500$ queries), while the best prior defense (i.e., adversarial training)
only attains $67.44\%$. Since AAA attacks SQA's general greedy strategy, such
advantages of AAA over 8 defenses can be consistently observed on 8
CIFAR-10/ImageNet models under 6 SQAs, using different attack targets and
bounds. Moreover, AAA calibrates better without hurting the accuracy. Our code
would be released. |
4,069 | Inference of a Rumor's Source in the Independent Cascade Model | We consider the so-called Independent Cascade Model for rumor spreading or
epidemic processes popularized by Kempe et al.\ [2003]. In this model, a small
subset of nodes from a network are the source of a rumor. In discrete time
steps, each informed node "infects" each of its uninformed neighbors with
probability $p$. While many facets of this process are studied in the
literature, less is known about the inference problem: given a number of
infected nodes in a network, can we learn the source of the rumor? In the
context of epidemiology this problem is often referred to as patient zero
problem. It belongs to a broader class of problems where the goal is to infer
parameters of the underlying spreading model, see, e.g., Lokhov [NeurIPS'16] or
Mastakouri et al. [NeurIPS'20].
In this work we present a maximum likelihood estimator for the rumor's
source, given a snapshot of the process in terms of a set of active nodes $X$
after $t$ steps. Our results show that, for cycle-free graphs, the likelihood
estimator undergoes a non-trivial phase transition as a function $t$. We
provide a rigorous analysis for two prominent classes of acyclic network,
namely $d$-regular trees and Galton-Watson trees, and verify empirically that
our heuristics work well in various general networks. |
4,070 | Phased Progressive Learning with Coupling-Regulation-Imbalance Loss for Imbalanced Classification | Deep neural networks generally perform poorly with datasets that suffer from
quantity imbalance and classification difficulty imbalance between different
classes. In order to alleviate the problem of dataset bias or domain shift in
the existing two-stage approaches, a phased progressive learning schedule was
proposed for smoothly transferring the training emphasis from representation
learning to upper classifier training. This has greater effectivity on datasets
that have more severe imbalances or smaller scales. A
coupling-regulation-imbalance loss function was designed, coupling a correction
term, Focal loss and LDAM loss. Coupling-regulation-imbalance loss can better
deal with quantity imbalance and outliers, while regulating focus-of-attention
of samples with a variety of classification difficulties. Excellent results
were achieved on multiple benchmark datasets using these approaches and they
can be easily generalized for other imbalanced classification models. Our code
will be open source soon. |
4,071 | Federated singular value decomposition for high dimensional data | Federated learning (FL) is emerging as a privacy-aware alternative to
classical cloud-based machine learning. In FL, the sensitive data remains in
data silos and only aggregated parameters are exchanged. Hospitals and research
institutions which are not willing to share their data can join a federated
study without breaching confidentiality. In addition to the extreme sensitivity
of biomedical data, the high dimensionality poses a challenge in the context of
federated genome-wide association studies (GWAS). In this article, we present a
federated singular value decomposition (SVD) algorithm, suitable for the
privacy-related and computational requirements of GWAS. Notably, the algorithm
has a transmission cost independent of the number of samples and is only weakly
dependent on the number of features, because the singular vectors associated
with the samples are never exchanged and the vectors associated with the
features only for a fixed number of iterations. Although motivated by GWAS, the
algorithm is generically applicable for both horizontally and vertically
partitioned data. |
4,072 | Empirical Phase Diagram for Three-layer Neural Networks with Infinite Width | Substantial work indicates that the dynamics of neural networks (NNs) is
closely related to their initialization of parameters. Inspired by the phase
diagram for two-layer ReLU NNs with infinite width (Luo et al., 2021), we make
a step towards drawing a phase diagram for three-layer ReLU NNs with infinite
width. First, we derive a normalized gradient flow for three-layer ReLU NNs and
obtain two key independent quantities to distinguish different dynamical
regimes for common initialization methods. With carefully designed experiments
and a large computation cost, for both synthetic datasets and real datasets, we
find that the dynamics of each layer also could be divided into a linear regime
and a condensed regime, separated by a critical regime. The criteria is the
relative change of input weights (the input weight of a hidden neuron consists
of the weight from its input layer to the hidden neuron and its bias term) as
the width approaches infinity during the training, which tends to $0$,
$+\infty$ and $O(1)$, respectively. In addition, we also demonstrate that
different layers can lie in different dynamical regimes in a training process
within a deep NN. In the condensed regime, we also observe the condensation of
weights in isolated orientations with low complexity. Through experiments under
three-layer condition, our phase diagram suggests a complicated dynamical
regimes consisting of three possible regimes, together with their mixture, for
deep NNs and provides a guidance for studying deep NNs in different
initialization regimes, which reveals the possibility of completely different
dynamics emerging within a deep NN for its different layers. |
4,073 | Associative Learning Mechanism for Drug-Target Interaction Prediction | As a necessary process in drug development, finding a drug compound that can
selectively bind to a specific protein is highly challenging and costly.
Drug-target affinity (DTA), which represents the strength of drug-target
interaction (DTI), has played an important role in the DTI prediction task over
the past decade. Although deep learning has been applied to DTA-related
research, existing solutions ignore fundamental correlations between molecular
substructures in molecular representation learning of drug compound
molecules/protein targets. Moreover, traditional methods lack the
interpretability of the DTA prediction process. This results in missing feature
information of intermolecular interactions, thereby affecting prediction
performance. Therefore, this paper proposes a DTA prediction method with
interactive learning and an autoencoder mechanism. The proposed model enhances
the corresponding ability to capture the feature information of a single
molecular sequence by the drug/protein molecular representation learning module
and supplements the information interaction between molecular sequence pairs by
the interactive information learning module. The DTA value prediction module
fuses the drug-target pair interaction information to output the predicted
value of DTA. Additionally, this paper theoretically proves that the proposed
method maximizes evidence lower bound (ELBO) for the joint distribution of the
DTA prediction model, which enhances the consistency of the probability
distribution between the actual value and the predicted value. The experimental
results confirm mutual transformer-drug target affinity (MT-DTA) achieves
better performance than other comparative methods. |
4,074 | DNNAbacus: Toward Accurate Computational Cost Prediction for Deep Neural Networks | Deep learning is attracting interest across a variety of domains, including
natural language processing, speech recognition, and computer vision. However,
model training is time-consuming and requires huge computational resources.
Existing works on the performance prediction of deep neural networks, which
mostly focus on the training time prediction of a few models, rely on
analytical models and result in high relative errors. %Optimizing task
scheduling and reducing job failures in data centers are essential to improve
resource utilization and reduce carbon emissions. This paper investigates the
computational resource demands of 29 classical deep neural networks and builds
accurate models for predicting computational costs. We first analyze the
profiling results of typical networks and demonstrate that the computational
resource demands of models with different inputs and hyperparameters are not
obvious and intuitive. We then propose a lightweight prediction approach
DNNAbacus with a novel network structural matrix for network representation.
DNNAbacus can accurately predict both memory and time cost for PyTorch and
TensorFlow models, which is also generalized to different hardware
architectures and can have zero-shot capability for unseen networks. Our
experimental results show that the mean relative error (MRE) is 0.9% with
respect to time and 2.8% with respect to memory for 29 classic models, which is
much lower than the state-of-the-art works. |
4,075 | Bias Discovery in Machine Learning Models for Mental Health | Fairness and bias are crucial concepts in artificial intelligence, yet they
are relatively ignored in machine learning applications in clinical psychiatry.
We computed fairness metrics and present bias mitigation strategies using a
model trained on clinical mental health data. We collected structured data
related to the admission, diagnosis, and treatment of patients in the
psychiatry department of the University Medical Center Utrecht. We trained a
machine learning model to predict future administrations of benzodiazepines on
the basis of past data. We found that gender plays an unexpected role in the
predictions-this constitutes bias. Using the AI Fairness 360 package, we
implemented reweighing and discrimination-aware regularization as bias
mitigation strategies, and we explored their implications for model
performance. This is the first application of bias exploration and mitigation
in a machine learning model trained on real clinical psychiatry data. |
4,076 | Optimality Conditions and Algorithms for Top-K Arm Identification | We consider the top-k arm identification problem for multi-armed bandits with
rewards belonging to a one-parameter canonical exponential family. The
objective is to select the set of k arms with the highest mean rewards by
sequential allocation of sampling efforts. We propose a unified optimal
allocation problem that identifies the complexity measures of this problem
under the fixed-confidence, fixed-budget settings, and the posterior
convergence rate from the Bayesian perspective. We provide the first
characterization of its optimality. We provide the first provably optimal
algorithm in the fixed-confidence setting for k>1. We also propose an efficient
heuristic algorithm for the top-k arm identification problem. Extensive
numerical experiments demonstrate superior performance compare to existing
methods in all three settings. |
4,077 | Ensemble Multi-Relational Graph Neural Networks | It is well established that graph neural networks (GNNs) can be interpreted
and designed from the perspective of optimization objective. With this clear
optimization objective, the deduced GNNs architecture has sound theoretical
foundation, which is able to flexibly remedy the weakness of GNNs. However,
this optimization objective is only proved for GNNs with single-relational
graph. Can we infer a new type of GNNs for multi-relational graphs by extending
this optimization objective, so as to simultaneously solve the issues in
previous multi-relational GNNs, e.g., over-parameterization? In this paper, we
propose a novel ensemble multi-relational GNNs by designing an ensemble
multi-relational (EMR) optimization objective. This EMR optimization objective
is able to derive an iterative updating rule, which can be formalized as an
ensemble message passing (EnMP) layer with multi-relations. We further analyze
the nice properties of EnMP layer, e.g., the relationship with multi-relational
personalized PageRank. Finally, a new multi-relational GNNs which well
alleviate the over-smoothing and over-parameterization issues are proposed.
Extensive experiments conducted on four benchmark datasets well demonstrate the
effectiveness of the proposed model. |
4,078 | Deep Reinforcement Learning for Multi-class Imbalanced Training | With the rapid growth of memory and computing power, datasets are becoming
increasingly complex and imbalanced. This is especially severe in the context
of clinical data, where there may be one rare event for many cases in the
majority class. We introduce an imbalanced classification framework, based on
reinforcement learning, for training extremely imbalanced data sets, and extend
it for use in multi-class settings. We combine dueling and double deep
Q-learning architectures, and formulate a custom reward function and
episode-training procedure, specifically with the added capability of handling
multi-class imbalanced training. Using real-world clinical case studies, we
demonstrate that our proposed framework outperforms current state-of-the-art
imbalanced learning methods, achieving more fair and balanced classification,
while also significantly improving the prediction of minority classes. |
4,079 | Boosting Tail Neural Network for Realtime Custom Keyword Spotting | In this paper, we propose a Boosting Tail Neural Network (BTNN) for improving
the performance of Realtime Custom Keyword Spotting (RCKS) that is still an
industrial challenge for demanding powerful classification ability with limited
computation resources. Inspired by Brain Science that a brain is only partly
activated for a nerve simulation and numerous machine learning algorithms are
developed to use a batch of weak classifiers to resolve arduous problems, which
are often proved to be effective. We show that this method is helpful to the
RCKS problem. The proposed approach achieve better performances in terms of
wakeup rate and false alarm.
In our experiments compared with those traditional algorithms that use only
one strong classifier, it gets 18\% relative improvement. We also point out
that this approach may be promising in future ASR exploration. |
4,080 | On statistic alignment for domain adaptation in structural health monitoring | The practical application of structural health monitoring (SHM) is often
limited by the availability of labelled data. Transfer learning - specifically
in the form of domain adaptation (DA) - gives rise to the possibility of
leveraging information from a population of physical or numerical structures,
by inferring a mapping that aligns the feature spaces. Typical DA methods rely
on nonparametric distance metrics, which require sufficient data to perform
density estimation. In addition, these methods can be prone to performance
degradation under class imbalance. To address these issues, statistic alignment
(SA) is discussed, with a demonstration of how these methods can be made robust
to class imbalance, including a special case of class imbalance called a
partial DA scenario. SA is demonstrated to facilitate damage localisation with
no target labels in a numerical case study, outperforming other
state-of-the-art DA methods. It is then shown to be capable of aligning the
feature spaces of a real heterogeneous population, the Z24 and KW51 bridges,
with only 220 samples used from the KW51 bridge. Finally, in scenarios where
more complex mappings are required for knowledge transfer, SA is shown to be a
vital pre-processing tool, increasing the performance of established DA
methods. |
4,081 | FedEntropy: Efficient Device Grouping for Federated Learning Using Maximum Entropy Judgment | Along with the popularity of Artificial Intelligence (AI) and
Internet-of-Things (IoT), Federated Learning (FL) has attracted steadily
increasing attentions as a promising distributed machine learning paradigm,
which enables the training of a central model on for numerous decentralized
devices without exposing their privacy. However, due to the biased data
distributions on involved devices, FL inherently suffers from low
classification accuracy in non-IID scenarios. Although various device grouping
method have been proposed to address this problem, most of them neglect both i)
distinct data distribution characteristics of heterogeneous devices, and ii)
contributions and hazards of local models, which are extremely important in
determining the quality of global model aggregation. In this paper, we present
an effective FL method named FedEntropy with a novel dynamic device grouping
scheme, which makes full use of the above two factors based on our proposed
maximum entropy judgement heuristic.Unlike existing FL methods that directly
aggregate local models returned from all the selected devices, in one FL round
FedEntropy firstly makes a judgement based on the pre-collected soft labels of
selected devices and then only aggregates the local models that can maximize
the overall entropy of these soft labels. Without collecting local models that
are harmful for aggregation, FedEntropy can effectively improve global model
accuracy while reducing the overall communication overhead. Comprehensive
experimental results on well-known benchmarks show that, FedEntropy not only
outperforms state-of-the-art FL methods in terms of model accuracy and
communication overhead, but also can be integrated into them to enhance their
classification performance. |
4,082 | Improving Human Image Synthesis with Residual Fast Fourier Transformation and Wasserstein Distance | With the rapid development of the Metaverse, virtual humans have emerged, and
human image synthesis and editing techniques, such as pose transfer, have
recently become popular. Most of the existing techniques rely on GANs, which
can generate good human images even with large variants and occlusions. But
from our best knowledge, the existing state-of-the-art method still has the
following problems: the first is that the rendering effect of the synthetic
image is not realistic, such as poor rendering of some regions. And the second
is that the training of GAN is unstable and slow to converge, such as model
collapse. Based on the above two problems, we propose several methods to solve
them. To improve the rendering effect, we use the Residual Fast Fourier
Transform Block to replace the traditional Residual Block. Then, spectral
normalization and Wasserstein distance are used to improve the speed and
stability of GAN training. Experiments demonstrate that the methods we offer
are effective at solving the problems listed above, and we get state-of-the-art
scores in LPIPS and PSNR. |
4,083 | PatchNR: Learning from Small Data by Patch Normalizing Flow Regularization | Learning neural networks using only a small amount of data is an important
ongoing research topic with tremendous potential for applications. In this
paper, we introduce a regularizer for the variational modeling of inverse
problems in imaging based on normalizing flows. Our regularizer, called
patchNR, involves a normalizing flow learned on patches of very few images. The
subsequent reconstruction method is completely unsupervised and the same
regularizer can be used for different forward operators acting on the same
class of images. By investigating the distribution of patches versus those of
the whole image class, we prove that our variational model is indeed a MAP
approach. Our model can be generalized to conditional patchNRs, if additional
supervised information is available. Numerical examples for low-dose CT,
limited-angle CT and superresolution of material images demonstrate that our
method provides high quality results among unsupervised methods, but requires
only few data. |
4,084 | Concurrent Credit Assignment for Data-efficient Reinforcement Learning | The capability to widely sample the state and action spaces is a key
ingredient toward building effective reinforcement learning algorithms. The
variational optimization principles exposed in this paper emphasize the
importance of an occupancy model to synthesizes the general distribution of the
agent's environmental states over which it can act (defining a virtual
``territory''). The occupancy model is the subject of frequent updates as the
exploration progresses and that new states are undisclosed during the course of
the training. By making a uniform prior assumption, the resulting objective
expresses a balance between two concurrent tendencies, namely the widening of
the occupancy space and the maximization of the rewards, reminding of the
classical exploration/exploitation trade-off. Implemented on an actor-critic
off-policy on classic continuous action benchmarks, it is shown to provide
significant increase in the sampling efficacy, that is reflected in a reduced
training time and higher returns, in both the dense and the sparse rewards
cases. |
4,085 | Naive Few-Shot Learning: Sequence Consistency Evaluation | Cognitive psychologists often use the term $\textit{fluid intelligence}$ to
describe the ability of humans to solve novel tasks without any prior training.
In contrast to humans, deep neural networks can perform cognitive tasks only
after extensive (pre-)training with a large number of relevant examples.
Motivated by fluid intelligence research in the cognitive sciences, we built a
benchmark task which we call sequence consistency evaluation (SCE) that can be
used to address this gap. Solving the SCE task requires the ability to extract
simple rules from sequences, a basic computation that is required for solving
various intelligence tests in humans. We tested $\textit{untrained}$ (naive)
deep learning models in the SCE task. Specifically, we compared Relation
Networks (RN) and Contrastive Predictive Coding (CPC), two models that can
extract simple rules from sequences, and found that the latter, which imposes a
structure on the predictable rule does better. We further found that simple
networks fare better in this task than complex ones. Finally, we show that this
approach can be used for security camera anomaly detection without any prior
training. |
4,086 | Graph Convolutional Reinforcement Learning for Collaborative Queuing Agents | In this paper, we explore the use of multi-agent deep learning as well as
learning to cooperate principles to meet stringent service level agreements, in
terms of throughput and end-to-end delay, for a set of classified network
flows. We consider agents built on top of a weighted fair queuing algorithm
that continuously set weights for three flow groups: gold, silver, and bronze.
We rely on a novel graph-convolution based, multi-agent reinforcement learning
approach known as DGN. As benchmarks, we propose centralized and distributed
deep Q-network approaches and evaluate their performances in different network,
traffic, and routing scenarios, highlighting the effectiveness of our proposals
and the importance of agent cooperation. We show that our DGN-based approach
meets stringent throughput and delay requirements across all scenarios. |
4,087 | Highly Accurate FMRI ADHD Classification using time distributed multi modal 3D CNNs | This work proposes an algorithm for fMRI data analysis for the classification
of ADHD disorders. There have been several breakthroughs in the analysis of
fMRI via 3D convolutional neural networks (CNNs). With these new techniques it
is possible to preserve the 3D spatial data of fMRI data. Additionally there
have been recent advances in the use of 3D generative adversarial neural
networks (GANs) for the generation of normal MRI data. This work utilizes multi
modal 3D CNNs with data augmentation from 3D GAN for ADHD prediction from fMRI.
By leveraging a 3D-GAN it would be possible to use deepfake data to enhance the
accuracy of 3D CNN classification of brain disorders. A comparison will be made
between a time distributed single modal 3D CNN model for classification and the
modified multi modal model with MRI data as well. |
4,088 | Learning Stabilizing Policies in Stochastic Control Systems | In this work, we address the problem of learning provably stable neural
network policies for stochastic control systems. While recent work has
demonstrated the feasibility of certifying given policies using martingale
theory, the problem of how to learn such policies is little explored. Here, we
study the effectiveness of jointly learning a policy together with a martingale
certificate that proves its stability using a single learning algorithm. We
observe that the joint optimization problem becomes easily stuck in local
minima when starting from a randomly initialized policy. Our results suggest
that some form of pre-training of the policy is required for the joint
optimization to repair and verify the policy successfully. |
4,089 | Realization Theory Of Recurrent Neural ODEs Using Polynomial System Embeddings | In this paper we show that neural ODE analogs of recurrent (ODE-RNN) and Long
Short-Term Memory (ODE-LSTM) networks can be algorithmically embeddeded into
the class of polynomial systems. This embedding preserves input-output behavior
and can suitably be extended to other neural DE architectures. We then use
realization theory of polynomial systems to provide necessary conditions for an
input-output map to be realizable by an ODE-LSTM and sufficient conditions for
minimality of such systems. These results represent the first steps towards
realization theory of recurrent neural ODE architectures, which is is expected
be useful for model reduction and learning algorithm analysis of recurrent
neural ODEs. |
4,090 | The Data-Production Dispositif | Machine learning (ML) depends on data to train and verify models. Very often,
organizations outsource processes related to data work (i.e., generating and
annotating data and evaluating outputs) through business process outsourcing
(BPO) companies and crowdsourcing platforms. This paper investigates outsourced
ML data work in Latin America by studying three platforms in Venezuela and a
BPO in Argentina. We lean on the Foucauldian notion of dispositif to define the
data-production dispositif as an ensemble of discourses, actions, and objects
strategically disposed to (re)produce power/knowledge relations in data and
labor. Our dispositif analysis comprises the examination of 210 data work
instruction documents, 55 interviews with data workers, managers, and
requesters, and participant observation. Our findings show that discourses
encoded in instructions reproduce and normalize the worldviews of requesters.
Precarious working conditions and economic dependency alienate workers, making
them obedient to instructions. Furthermore, discourses and social contexts
materialize in artifacts, such as interfaces and performance metrics, limiting
workers' agency and normalizing specific ways of interpreting data. We conclude
by stressing the importance of counteracting the data-production dispositif by
fighting alienation and precarization, and empowering data workers to become
assets in the quest for high-quality data. |
4,091 | Bandwidth Selection for Gaussian Kernel Ridge Regression via Jacobian Control | Most machine learning methods depend on the tuning of hyper-parameters. For
kernel ridge regression (KRR) with the Gaussian kernel, the hyper-parameter is
the bandwidth. The bandwidth specifies the length-scale of the kernel and has
to be carefully selected in order to obtain a model with good generalization.
The default method for bandwidth selection is cross-validation, which often
yields good results, albeit at high computational costs. Furthermore, the
estimates provided by cross-validation tend to have very high variance,
especially when training data are scarce. Inspired by Jacobian regularization,
we formulate how the derivatives of the functions inferred by KRR with the
Gaussian kernel depend on the kernel bandwidth. We then use this expression to
propose a closed-form, computationally feather-light, bandwidth selection
method based on controlling the Jacobian. In addition, the Jacobian expression
illuminates how the bandwidth selection is a trade-off between the smoothness
of the inferred function, and the conditioning of the training data kernel
matrix. We show on real and synthetic data that compared to cross-validation,
our method is considerably more stable in terms of bandwidth selection, and,
for small data sets, provides better predictions. |
4,092 | 3D helical CT reconstruction with memory efficient invertible Learned Primal-Dual method | Helical acquisition geometry is the most common geometry used in computed
tomography (CT) scanners for medical imaging. We adapt the invertible Learned
Primal-Dual (iLPD) deep neural network architecture so that it can be applied
to helical 3D CT reconstruction. We achieve this by splitting the geometry and
the data in parts that fit the memory and by splitting images into
corresponding sub-volumes. The architecture can be applied to images different
in size along the rotation axis. We perform the experiments on tomographic data
simulated from realistic helical geometries. |
4,093 | Causal Influences Decouple From Their Underlying Network Structure In Echo State Networks | Echo State Networks (ESN) are versatile recurrent neural network models in
which the hidden layer remains unaltered during training. Interactions among
nodes of this static backbone produce diverse representations of the given
stimuli that are harnessed by a read-out mechanism to perform computations
needed for solving a given task. ESNs are accessible models of neuronal
circuits, since they are relatively inexpensive to train. Therefore, ESNs have
become attractive for neuroscientists studying the relationship between neural
structure, function, and behavior. For instance, it is not yet clear how
distinctive connectivity patterns of brain networks support effective
interactions among their nodes and how these patterns of interactions give rise
to computation. To address this question, we employed an ESN with a
biologically inspired structure and used a systematic multi-site lesioning
framework to quantify the causal contribution of each node to the network's
output, thus providing a causal link between network structure and behavior. We
then focused on the structure-function relationship and decomposed the causal
influence of each node on all other nodes, using the same lesioning framework.
We found that nodes in a properly engineered ESN interact largely irrespective
of the network's underlying structure. However, in a network with the same
topology and a non-optimal parameter set, the underlying connectivity patterns
determine the node interactions. Our results suggest that causal
structure-function relations in ESNs can be decomposed into two components,
direct and indirect interactions. The former are based on influences relying on
structural connections. The latter describe the effective communication between
any two nodes through other intermediate nodes. These widely distributed
indirect interactions may crucially contribute to the efficient performance of
ESNs. |
4,094 | Assessing the Quality of Computational Notebooks for a Frictionless Transition from Exploration to Production | The massive trend of integrating data-driven AI capabilities into traditional
software systems is rising new intriguing challenges. One of such challenges is
achieving a smooth transition from the explorative phase of Machine Learning
projects - in which data scientists build prototypical models in the lab - to
their production phase - in which software engineers translate prototypes into
production-ready AI components. To narrow down the gap between these two
phases, tools and practices adopted by data scientists might be improved by
incorporating consolidated software engineering solutions. In particular,
computational notebooks have a prominent role in determining the quality of
data science prototypes. In my research project, I address this challenge by
studying the best practices for collaboration with computational notebooks and
proposing proof-of-concept tools to foster guidelines compliance. |
4,095 | Pynblint: a Static Analyzer for Python Jupyter Notebooks | Jupyter Notebook is the tool of choice of many data scientists in the early
stages of ML workflows. The notebook format, however, has been criticized for
inducing bad programming practices; indeed, researchers have already shown that
open-source repositories are inundated by poor-quality notebooks. Low-quality
output from the prototypical stages of ML workflows constitutes a clear
bottleneck towards the productization of ML models. To foster the creation of
better notebooks, we developed Pynblint, a static analyzer for Jupyter
notebooks written in Python. The tool checks the compliance of notebooks (and
surrounding repositories) with a set of empirically validated best practices
and provides targeted recommendations when violations are detected. |
4,096 | How Human is Human Evaluation? Improving the Gold Standard for NLG with Utility Theory | Human ratings are treated as the gold standard in NLG evaluation. The
standard protocol is to collect ratings of generated text, average across
annotators, and then rank NLG systems by their average scores. However, little
consideration has been given as to whether this approach faithfully captures
human preferences. In this work, we analyze this standard protocol through the
lens of utility theory in economics. We first identify the implicit assumptions
it makes about annotators and find that these assumptions are often violated in
practice, in which case annotator ratings become an unfaithful reflection of
their preferences. The most egregious violations come from using Likert scales,
which provably reverse the direction of the true preference in certain cases.
We suggest improvements to the standard protocol to make it more theoretically
sound, but even in its improved form, it cannot be used to evaluate open-ended
tasks like story generation. For the latter, we propose a new evaluation
protocol called $\textit{system-level probabilistic assessment}$ (SPA). In our
experiments, we find that according to SPA, annotators prefer larger GPT-3
variants to smaller ones -- as expected -- with all comparisons being
statistically significant. In contrast, the standard protocol only yields
significant results half the time. |
4,097 | Compression-aware Training of Neural Networks using Frank-Wolfe | Many existing Neural Network pruning approaches either rely on retraining to
compensate for pruning-caused performance degradation or they induce strong
biases to converge to a specific sparse solution throughout training. A third
paradigm obtains a wide range of compression ratios from a single dense
training run while also avoiding retraining. Recent work of Pokutta et al.
(2020) and Miao et al. (2022) suggests that the Stochastic Frank-Wolfe (SFW)
algorithm is particularly suited for training state-of-the-art models that are
robust to compression. We propose leveraging $k$-support norm ball constraints
and demonstrate significant improvements over the results of Miao et al. (2022)
in the case of unstructured pruning. We also extend these ideas to the
structured pruning domain and propose novel approaches to both ensure
robustness to the pruning of convolutional filters as well as to low-rank
tensor decompositions of convolutional layers. In the latter case, our approach
performs on-par with nuclear-norm regularization baselines while requiring only
half of the computational resources. Our findings also indicate that the
robustness of SFW-trained models largely depends on the gradient rescaling of
the learning rate and we establish a theoretical foundation for that practice. |
4,098 | Large Language Models are Zero-Shot Reasoners | Pretrained large language models (LLMs) are widely used in many sub-fields of
natural language processing (NLP) and generally known as excellent few-shot
learners with task-specific exemplars. Notably, chain of thought (CoT)
prompting, a recent technique for eliciting complex multi-step reasoning
through step-by-step answer examples, achieved the state-of-the-art
performances in arithmetics and symbolic reasoning, difficult system-2 tasks
that do not follow the standard scaling laws for LLMs. While these successes
are often attributed to LLMs' ability for few-shot learning, we show that LLMs
are decent zero-shot reasoners by simply adding "Let's think step by step"
before each answer. Experimental results demonstrate that our Zero-shot-CoT,
using the same single prompt template, significantly outperforms zero-shot LLM
performances on diverse benchmark reasoning tasks including arithmetics
(MultiArith, GSM8K, AQUA-RAT, SVAMP), symbolic reasoning (Last Letter, Coin
Flip), and other logical reasoning tasks (Date Understanding, Tracking Shuffled
Objects), without any hand-crafted few-shot examples, e.g. increasing the
accuracy on MultiArith from 17.7% to 78.7% and GSM8K from 10.4% to 40.7% with
175B parameter InstructGPT model, as well as similar magnitudes of improvements
with another off-the-shelf large model, 540B parameter PaLM. The versatility of
this single prompt across very diverse reasoning tasks hints at untapped and
understudied fundamental zero-shot capabilities of LLMs, suggesting high-level,
multi-task broad cognitive capabilities may be extracted by simple prompting.
We hope our work not only serves as the minimal strongest zero-shot baseline
for the challenging reasoning benchmarks, but also highlights the importance of
carefully exploring and analyzing the enormous zero-shot knowledge hidden
inside LLMs before crafting finetuning datasets or few-shot exemplars. |
4,099 | An Adaptive Contrastive Learning Model for Spike Sorting | Brain-computer interfaces (BCIs), is ways for electronic devices to
communicate directly with the brain. For most medical-type brain-computer
interface tasks, the activity of multiple units of neurons or local field
potentials is sufficient for decoding. But for BCIs used in neuroscience
research, it is important to separate out the activity of individual neurons.
With the development of large-scale silicon technology and the increasing
number of probe channels, artificially interpreting and labeling spikes is
becoming increasingly impractical. In this paper, we propose a novel modeling
framework: Adaptive Contrastive Learning Model that learns representations from
spikes through contrastive learning based on the maximizing mutual information
loss function as a theoretical basis. Based on the fact that data with similar
features share the same labels whether they are multi-classified or
binary-classified. With this theoretical support, we simplify the
multi-classification problem into multiple binary-classification, improving
both the accuracy and the runtime efficiency. Moreover, we also introduce a
series of enhancements for the spikes, while solving the problem that the
classification effect is affected because of the overlapping spikes. |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.