Unnamed: 0.1
int64 0
113k
| Unnamed: 0
float64 0
113k
⌀ | title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|---|---|
106,000 | 106,000 |
What makes useful auxiliary tasks in reinforcement learning:
investigating the effect of the target policy
|
Auxiliary tasks have been argued to be useful for representation learning in
reinforcement learning. Although many auxiliary tasks have been empirically
shown to be effective for accelerating learning on the main task, it is not yet
clear what makes useful auxiliary tasks. Some of the most promising results are
on the pixel control, reward prediction, and the next state prediction
auxiliary tasks; however, the empirical results are mixed, showing substantial
improvements in some cases and marginal improvements in others. Careful
investigations of how auxiliary tasks help the learning of the main task is
necessary. In this paper, we take a step studying the effect of the target
policies on the usefulness of the auxiliary tasks formulated as general value
functions. General value functions consist of three core elements: 1) policy 2)
cumulant 3) continuation function. Our focus on the role of the target policy
of the auxiliary tasks is motivated by the fact that the target policy
determines the behavior about which the agent wants to make a prediction and
the state-action distribution that the agent is trained on, which further
affects the main task learning. Our study provides insights about questions
such as: Does a greedy policy result in bigger improvement gains compared to
other policies? Is it best to set the auxiliary task policy to be the same as
the main task policy? Does the choice of the target policy have a substantial
effect on the achieved performance gain or simple strategies for setting the
policy, such as using a uniformly random policy, work as well? Our empirical
results suggest that: 1) Auxiliary tasks with the greedy policy tend to be
useful. 2) Most policies, including a uniformly random policy, tend to improve
over the baseline. 3) Surprisingly, the main task policy tends to be less
useful compared to other policies.
|
106,001 | 106,001 |
Connect, Not Collapse: Explaining Contrastive Learning for Unsupervised
Domain Adaptation
|
We consider unsupervised domain adaptation (UDA), where labeled data from a
source domain (e.g., photographs) and unlabeled data from a target domain
(e.g., sketches) are used to learn a classifier for the target domain.
Conventional UDA methods (e.g., domain adversarial training) learn
domain-invariant features to improve generalization to the target domain. In
this paper, we show that contrastive pre-training, which learns features on
unlabeled source and target data and then fine-tunes on labeled source data, is
competitive with strong UDA methods. However, we find that contrastive
pre-training does not learn domain-invariant features, diverging from
conventional UDA intuitions. We show theoretically that contrastive
pre-training can learn features that vary subtantially across domains but still
generalize to the target domain, by disentangling domain and class information.
Our results suggest that domain invariance is not necessary for UDA. We
empirically validate our theory on benchmark vision datasets.
|
106,002 | 106,002 |
Robust and Efficient Aggregation for Distributed Learning
|
Distributed learning paradigms, such as federated and decentralized learning,
allow for the coordination of models across a collection of agents, and without
the need to exchange raw data. Instead, agents compute model updates locally
based on their available data, and subsequently share the update model with a
parameter server or their peers. This is followed by an aggregation step, which
traditionally takes the form of a (weighted) average. Distributed learning
schemes based on averaging are known to be susceptible to outliers. A single
malicious agent is able to drive an averaging-based distributed learning
algorithm to an arbitrarily poor model. This has motivated the development of
robust aggregation schemes, which are based on variations of the median and
trimmed mean. While such procedures ensure robustness to outliers and malicious
behavior, they come at the cost of significantly reduced sample efficiency.
This means that current robust aggregation schemes require significantly higher
agent participation rates to achieve a given level of performance than their
mean-based counterparts in non-contaminated settings. In this work we remedy
this drawback by developing statistically efficient and robust aggregation
schemes for distributed learning.
|
106,003 | 106,003 |
Monarch: Expressive Structured Matrices for Efficient and Accurate
Training
|
Large neural networks excel in many domains, but they are expensive to train
and fine-tune. A popular approach to reduce their compute or memory
requirements is to replace dense weight matrices with structured ones (e.g.,
sparse, low-rank, Fourier transform). These methods have not seen widespread
adoption (1) in end-to-end training due to unfavorable efficiency--quality
tradeoffs, and (2) in dense-to-sparse fine-tuning due to lack of tractable
algorithms to approximate a given dense weight matrix. To address these issues,
we propose a class of matrices (Monarch) that is hardware-efficient (they are
parameterized as products of two block-diagonal matrices for better hardware
utilization) and expressive (they can represent many commonly used transforms).
Surprisingly, the problem of approximating a dense weight matrix with a Monarch
matrix, though nonconvex, has an analytical optimal solution. These properties
of Monarch matrices unlock new ways to train and fine-tune sparse and dense
models. We empirically validate that Monarch can achieve favorable
accuracy-efficiency tradeoffs in several end-to-end sparse training
applications: speeding up ViT and GPT-2 training on ImageNet classification and
Wikitext-103 language modeling by 2x with comparable model quality, and
reducing the error on PDE solving and MRI reconstruction tasks by 40%. In
sparse-to-dense training, with a simple technique called "reverse
sparsification," Monarch matrices serve as a useful intermediate representation
to speed up GPT-2 pretraining on OpenWebText by 2x without quality drop. The
same technique brings 23% faster BERT pretraining than even the very optimized
implementation from Nvidia that set the MLPerf 1.1 record. In dense-to-sparse
fine-tuning, as a proof-of-concept, our Monarch approximation algorithm speeds
up BERT fine-tuning on GLUE by 1.7x with comparable accuracy.
|
106,004 | 106,004 |
Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language
|
Large pretrained (e.g., "foundation") models exhibit distinct capabilities
depending on the domain of data they are trained on. While these domains are
generic, they may only barely overlap. For example, visual-language models
(VLMs) are trained on Internet-scale image captions, but large language models
(LMs) are further trained on Internet-scale text with no images (e.g.,
spreadsheets, SAT questions, code). As a result, these models store different
forms of commonsense knowledge across different domains. In this work, we show
that this diversity is symbiotic, and can be leveraged through Socratic Models
(SMs): a modular framework in which multiple pretrained models may be composed
zero-shot i.e., via multimodal-informed prompting, to exchange information with
each other and capture new multimodal capabilities, without requiring
finetuning. With minimal engineering, SMs are not only competitive with
state-of-the-art zero-shot image captioning and video-to-text retrieval, but
also enable new applications such as (i) answering free-form questions about
egocentric video, (ii) engaging in multimodal assistive dialogue with people
(e.g., for cooking recipes) by interfacing with external APIs and databases
(e.g., web search), and (iii) robot perception and planning.
|
106,005 | 106,005 |
From Statistical to Causal Learning
|
We describe basic ideas underlying research to build and understand
artificially intelligent systems: from symbolic approaches via statistical
learning to interventional models relying on concepts of causality. Some of the
hard open problems of machine learning and AI are intrinsically related to
causality, and progress may require advances in our understanding of how to
model and infer causality from data.
|
106,006 | 106,006 |
Learning the conditional law: signatures and conditional GANs in
filtering and prediction of diffusion processes
|
We consider the filtering and prediction problem for a diffusion process. The
signal and observation are modeled by stochastic differential equations (SDEs)
driven by Wiener processes. In classical estimation theory, measure-valued
stochastic partial differential equations (SPDEs) are derived for the filtering
and prediction measures. These equations can be hard to solve numerically. We
provide an approximation algorithm using conditional generative adversarial
networks (GANs) and signatures, an object from rough path theory. The signature
of a sufficiently smooth path determines the path completely. In some cases,
GANs based on signatures have been shown to efficiently approximate the law of
a stochastic process. In this paper we extend this method to approximate the
prediction measure conditional to noisy observation. We use controlled
differential equations (CDEs) as universal approximators to propose an
estimator for the conditional and prediction law. We show well-posedness in
providing a rigorous mathematical framework. Numerical results show the
efficiency of our algorithm.
|
106,007 | 106,007 |
On the Importance of Asymmetry for Siamese Representation Learning
|
Many recent self-supervised frameworks for visual representation learning are
based on certain forms of Siamese networks. Such networks are conceptually
symmetric with two parallel encoders, but often practically asymmetric as
numerous mechanisms are devised to break the symmetry. In this work, we conduct
a formal study on the importance of asymmetry by explicitly distinguishing the
two encoders within the network -- one produces source encodings and the other
targets. Our key insight is keeping a relatively lower variance in target than
source generally benefits learning. This is empirically justified by our
results from five case studies covering different variance-oriented designs,
and is aligned with our preliminary theoretical analysis on the baseline.
Moreover, we find the improvements from asymmetric designs generalize well to
longer training schedules, multiple other frameworks and newer backbones.
Finally, the combined effect of several asymmetric designs achieves a
state-of-the-art accuracy on ImageNet linear probing and competitive results on
downstream transfer. We hope our exploration will inspire more research in
exploiting asymmetry for Siamese representation learning.
|
106,008 | 106,008 |
Simplicial Embeddings in Self-Supervised Learning and Downstream
Classification
|
We introduce Simplicial Embeddings (SEMs) as a way to constrain the encoded
representations of a self-supervised model to $L$ simplices of $V$ dimensions
each using a Softmax operation. This procedure imposes a structure on the
representations that reduce their expressivity for training downstream
classifiers, which helps them generalize better. Specifically, we show that the
temperature $\tau$ of the Softmax operation controls for the SEM
representation's expressivity, allowing us to derive a tighter downstream
classifier generalization bound than that for classifiers using unnormalized
representations. We empirically demonstrate that SEMs considerably improve
generalization on natural image datasets such as CIFAR-100 and ImageNet.
Finally, we also present evidence of the emergence of semantically relevant
features in SEMs, a pattern that is absent from baseline self-supervised
models.
|
106,009 | 106,009 |
Visual explanations for polyp detection: How medical doctors assess
intrinsic versus extrinsic explanations
|
Deep learning has in recent years achieved immense success in all areas of
computer vision and has the potential of assisting medical doctors in analyzing
visual content for disease and other abnormalities. However, the current state
of deep learning is very much a black box, making medical professionals highly
skeptical about integrating these methods into clinical practice. Several
methods have been proposed in order to shine some light onto these black boxes,
but there is no consensus on the opinion of the medical doctors that will
consume these explanations. This paper presents a study asking medical doctors
about their opinion of current state-of-the-art explainable artificial
intelligence methods when applied to a gastrointestinal disease detection use
case. We compare two different categories of explanation methods, intrinsic and
extrinsic, and gauge their opinion of the current value of these explanations.
The results indicate that intrinsic explanations are preferred and that
explanation.
|
106,010 | 106,010 |
CogNGen: Constructing the Kernel of a Hyperdimensional Predictive
Processing Cognitive Architecture
|
We present a new cognitive architecture that combines two neurobiologically
plausible, computational models: (1) a variant of predictive processing known
as neural generative coding (NGC) and (2) hyperdimensional, vector-symbolic
models of human memory. We draw inspiration from well-known cognitive
architectures such as ACT-R, Soar, Leabra, and Spaun/Nengo. Our cognitive
architecture, the COGnitive Neural GENerative system (CogNGen), is in broad
agreement with these architectures, but provides a level of detail between
ACT-R's high-level, symbolic description of human cognition and Spaun's
low-level neurobiological description. CogNGen creates the groundwork for
developing agents that learn continually from diverse tasks and model human
performance at larger scales than what is possible with existent cognitive
architectures. We aim to develop a cognitive architecture that has the power of
modern machine learning techniques while retaining long-term memory,
single-trial learning, transfer-learning, planning, and other capacities
associated with high-level cognition. We test CogNGen on a set of maze-learning
tasks, including mazes that test short-term memory and planning, and find that
the addition of vector-symbolic models of memory improves the ability of the
NGC reinforcement learning model to master the maze task. Future work includes
testing CogNGen on more tasks and exploring methods for efficiently scaling
hyperdimensional memory models to lifetime learning.
|
106,011 | 106,011 |
Universal Lymph Node Detection in T2 MRI using Neural Networks
|
Purpose: Identification of abdominal Lymph Nodes (LN) that are suspicious for
metastasis in T2 Magnetic Resonance Imaging (MRI) scans is critical for staging
of lymphoproliferative diseases. Prior work on LN detection has been limited to
specific anatomical regions of the body (pelvis, rectum) in single MR slices.
Therefore, the development of a universal approach to detect LN in full T2 MRI
volumes is highly desirable.
Methods: In this study, a Computer Aided Detection (CAD) pipeline to
universally identify abdominal LN in volumetric T2 MRI using neural networks is
proposed. First, we trained various neural network models for detecting LN:
Faster RCNN with and without Hard Negative Example Mining (HNEM), FCOS,
FoveaBox, VFNet, and Detection Transformer (DETR). Next, we show that the
state-of-the-art (SOTA) VFNet model with Adaptive Training Sample Selection
(ATSS) outperforms Faster RCNN with HNEM. Finally, we ensembled models that
surpassed a 45% mAP threshold. We found that the VFNet model and one-stage
model ensemble can be interchangeably used in the CAD pipeline.
Results: Experiments on 122 test T2 MRI volumes revealed that VFNet achieved
a 51.1% mAP and 78.7% recall at 4 false positives (FP) per volume, while the
one-stage model ensemble achieved a mAP of 52.3% and sensitivity of 78.7% at
4FP.
Conclusion: Our contribution is a CAD pipeline that detects LN in T2 MRI
volumes, resulting in a sensitivity improvement of $\sim$14 points over the
current SOTA method for LN detection (sensitivity of 78.7% at 4 FP vs. 64.6% at
5 FP per volume).
|
106,012 | 106,012 |
Bayesian Image Super-Resolution with Deep Modeling of Image Statistics
|
Modeling statistics of image priors is useful for image super-resolution, but
little attention has been paid from the massive works of deep learning-based
methods. In this work, we propose a Bayesian image restoration framework, where
natural image statistics are modeled with the combination of smoothness and
sparsity priors. Concretely, firstly we consider an ideal image as the sum of a
smoothness component and a sparsity residual, and model real image degradation
including blurring, downscaling, and noise corruption. Then, we develop a
variational Bayesian approach to infer their posteriors. Finally, we implement
the variational approach for single image super-resolution (SISR) using deep
neural networks, and propose an unsupervised training strategy. The experiments
on three image restoration tasks, \textit{i.e.,} ideal SISR, realistic SISR,
and real-world SISR, demonstrate that our method has superior model
generalizability against varying noise levels and degradation kernels and is
effective in unsupervised SISR. The code and resulting models are released via
\url{https://zmiclab.github.io/projects.html}.
|
106,013 | 106,013 |
Explainable and Interpretable Diabetic Retinopathy Classification Based
on Neural-Symbolic Learning
|
In this paper, we propose an explainable and interpretable diabetic
retinopathy (ExplainDR) classification model based on neural-symbolic learning.
To gain explainability, a highlevel symbolic representation should be
considered in decision making. Specifically, we introduce a human-readable
symbolic representation, which follows a taxonomy style of diabetic retinopathy
characteristics related to eye health conditions to achieve explainability. We
then include humanreadable features obtained from the symbolic representation
in the disease prediction. Experimental results on a diabetic retinopathy
classification dataset show that our proposed ExplainDR method exhibits
promising performance when compared to that from state-of-the-art methods
applied to the IDRiD dataset, while also providing interpretability and
explainability.
|
106,014 | 106,014 |
Learning Neural Acoustic Fields
|
Our environment is filled with rich and dynamic acoustic information. When we
walk into a cathedral, the reverberations as much as appearance inform us of
the sanctuary's wide open space. Similarly, as an object moves around us, we
expect the sound emitted to also exhibit this movement. While recent advances
in learned implicit functions have led to increasingly higher quality
representations of the visual world, there have not been commensurate advances
in learning spatial auditory representations. To address this gap, we introduce
Neural Acoustic Fields (NAFs), an implicit representation that captures how
sounds propagate in a physical scene. By modeling acoustic propagation in a
scene as a linear time-invariant system, NAFs learn to continuously map all
emitter and listener location pairs to a neural impulse response function that
can then be applied to arbitrary sounds. We demonstrate that the continuous
nature of NAFs enables us to render spatial acoustics for a listener at an
arbitrary location, and can predict sound propagation at novel locations. We
further show that the representation learned by NAFs can help improve visual
learning with sparse views. Finally, we show that a representation informative
of scene structure emerges during the learning of NAFs.
|
106,015 | 106,015 |
TopTemp: Parsing Precipitate Structure from Temper Topology
|
Technological advances are in part enabled by the development of novel
manufacturing processes that give rise to new materials or material property
improvements. Development and evaluation of new manufacturing methodologies is
labor-, time-, and resource-intensive expensive due to complex, poorly defined
relationships between advanced manufacturing process parameters and the
resulting microstructures. In this work, we present a topological
representation of temper (heat-treatment) dependent material micro-structure,
as captured by scanning electron microscopy, called TopTemp. We show that this
topological representation is able to support temper classification of
microstructures in a data limited setting, generalizes well to previously
unseen samples, is robust to image perturbations, and captures domain
interpretable features. The presented work outperforms conventional deep
learning baselines and is a first step towards improving understanding of
process parameters and resulting material properties.
|
106,016 | 106,016 |
UNetFormer: A Unified Vision Transformer Model and Pre-Training
Framework for 3D Medical Image Segmentation
|
Vision Transformers (ViT)s have recently become popular due to their
outstanding modeling capabilities, in particular for capturing long-range
information, and scalability to dataset and model sizes which has led to
state-of-the-art performance in various computer vision and medical image
analysis tasks. In this work, we introduce a unified framework consisting of
two architectures, dubbed UNetFormer, with a 3D Swin Transformer-based encoder
and Convolutional Neural Network (CNN) and transformer-based decoders. In the
proposed model, the encoder is linked to the decoder via skip connections at
five different resolutions with deep supervision. The design of proposed
architecture allows for meeting a wide range of trade-off requirements between
accuracy and computational cost. In addition, we present a methodology for
self-supervised pre-training of the encoder backbone via learning to predict
randomly masked volumetric tokens using contextual information of visible
tokens. We pre-train our framework on a cohort of $5050$ CT images, gathered
from publicly available CT datasets, and present a systematic investigation of
various components such as masking ratio and patch size that affect the
representation learning capability and performance of downstream tasks. We
validate the effectiveness of our pre-training approach by fine-tuning and
testing our model on liver and liver tumor segmentation task using the Medical
Segmentation Decathlon (MSD) dataset and achieve state-of-the-art performance
in terms of various segmentation metrics. To demonstrate its generalizability,
we train and test the model on BraTS 21 dataset for brain tumor segmentation
using MRI images and outperform other methods in terms of Dice score. Code:
https://github.com/Project-MONAI/research-contributions
|
106,017 | 106,017 |
Application of Dimensional Reduction in Artificial Neural Networks to
Improve Emergency Department Triage During Chemical Mass Casualty Incidents
|
Chemical Mass Casualty Incidents (MCI) place a heavy burden on hospital staff
and resources. Machine Learning (ML) tools can provide efficient decision
support to caregivers. However, ML models require large volumes of data for the
most accurate results, which is typically not feasible in the chaotic nature of
a chemical MCI. This study examines the application of four statistical
dimension reduction techniques: Random Selection, Covariance/Variance,
Pearson's Linear Correlation, and Principle Component Analysis to reduce a
dataset of 311 hazardous chemicals and 79 related signs and symptoms (SSx). An
Artificial Neural Network pipeline was developed to create comparative models.
Results show that the number of signs and symptoms needed to determine a
chemical culprit can be reduced to nearly 40 SSx without losing significant
model accuracy. Evidence also suggests that the application of dimension
reduction methods can improve ANN model performance accuracy.
|
106,018 | 106,018 |
SIMBAR: Single Image-Based Scene Relighting For Effective Data
Augmentation For Automated Driving Vision Tasks
|
Real-world autonomous driving datasets comprise of images aggregated from
different drives on the road. The ability to relight captured scenes to unseen
lighting conditions, in a controllable manner, presents an opportunity to
augment datasets with a richer variety of lighting conditions, similar to what
would be encountered in the real-world. This paper presents a novel image-based
relighting pipeline, SIMBAR, that can work with a single image as input. To the
best of our knowledge, there is no prior work on scene relighting leveraging
explicit geometric representations from a single image. We present qualitative
comparisons with prior multi-view scene relighting baselines. To further
validate and effectively quantify the benefit of leveraging SIMBAR for data
augmentation for automated driving vision tasks, object detection and tracking
experiments are conducted with a state-of-the-art method, a Multiple Object
Tracking Accuracy (MOTA) of 93.3% is achieved with CenterTrack on
SIMBAR-augmented KITTI - an impressive 9.0% relative improvement over the
baseline MOTA of 85.6% with CenterTrack on original KITTI, both models trained
from scratch and tested on Virtual KITTI. For more details and SIMBAR relit
datasets, please visit our project website (https://simbarv1.github.io/).
|
106,019 | 106,019 |
Cluster-based ensemble learning for wind power modeling with
meteorological wind data
|
Optimal implementation and monitoring of wind energy generation hinge on
reliable power modeling that is vital for understanding turbine control, farm
operational optimization, and grid load balance. Based on the idea of similar
wind condition leads to similar wind power; this paper constructs a modeling
scheme that orderly integrates three types of ensemble learning algorithms,
bagging, boosting, and stacking, and clustering approaches to achieve optimal
power modeling. It also investigates applications of different clustering
algorithms and methodology for determining cluster numbers in wind power
modeling. The results reveal that all ensemble models with clustering exploit
the intrinsic information of wind data and thus outperform models without it by
approximately 15% on average. The model with the best farthest first clustering
is computationally rapid and performs exceptionally well with an improvement of
around 30%. The modeling is further boosted by about 5% by introducing stacking
that fuses ensembles with varying clusters. The proposed modeling framework
thus demonstrates promise by delivering efficient and robust modeling
performance.
|
106,020 | 106,020 |
Knowledge distillation with error-correcting transfer learning for wind
power prediction
|
Wind power prediction, especially for turbines, is vital for the operation,
controllability, and economy of electricity companies. Hybrid methodologies
combining advanced data science with weather forecasting have been
incrementally applied to the predictions. Nevertheless, individually modeling
massive turbines from scratch and downscaling weather forecasts to turbine size
are neither easy nor economical. Aiming at it, this paper proposes a novel
framework with mathematical underpinnings for turbine power prediction. This
framework is the first time to incorporate knowledge distillation into energy
forecasting, enabling accurate and economical constructions of turbine models
by learning knowledge from the well-established park model. Besides, park-scale
weather forecasts non-explicitly are mapped to turbines by transfer learning of
predicted power errors, achieving model correction for better performance. The
proposed framework is deployed on five turbines featuring various terrains in
an Arctic wind park, the results are evaluated against the competitors of
ablation investigation. The major findings reveal that the proposed framework,
developed on favorable knowledge distillation and transfer learning parameters
tuning, yields performance boosts from 3.3 % to 23.9 % over its competitors.
This advantage also exists in terms of wind energy physics and computing
efficiency, which are verified by the prediction quality rate and calculation
time.
|
106,021 | 106,021 |
Hysteresis-Based RL: Robustifying Reinforcement Learning-based Control
Policies via Hybrid Control
|
Reinforcement learning (RL) is a promising approach for deriving control
policies for complex systems. As we show in two control problems, the derived
policies from using the Proximal Policy Optimization (PPO) and Deep Q-Network
(DQN) algorithms may lack robustness guarantees. Motivated by these issues, we
propose a new hybrid algorithm, which we call Hysteresis-Based RL (HyRL),
augmenting an existing RL algorithm with hysteresis switching and two stages of
learning. We illustrate its properties in two examples for which PPO and DQN
fail.
|
106,022 | 106,022 |
Learnable latent embeddings for joint behavioral and neural analysis
|
Mapping behavioral actions to neural activity is a fundamental goal of
neuroscience. As our ability to record large neural and behavioral data
increases, there is growing interest in modeling neural dynamics during
adaptive behaviors to probe neural representations. In particular, neural
latent embeddings can reveal underlying correlates of behavior, yet, we lack
non-linear techniques that can explicitly and flexibly leverage joint behavior
and neural data. Here, we fill this gap with a novel method, CEBRA, that
jointly uses behavioral and neural data in a hypothesis- or discovery-driven
manner to produce consistent, high-performance latent spaces. We validate its
accuracy and demonstrate our tool's utility for both calcium and
electrophysiology datasets, across sensory and motor tasks, and in simple or
complex behaviors across species. It allows for single and multi-session
datasets to be leveraged for hypothesis testing or can be used label-free.
Lastly, we show that CEBRA can be used for the mapping of space, uncovering
complex kinematic features, and rapid, high-accuracy decoding of natural movies
from visual cortex.
|
106,023 | 106,023 |
Assimilation of Satellite Active Fires Data
|
Wildland fires pose an increasingly serious problem in our society. The
number and severity of these fires has been rising for many years. Wildfires
pose direct threats to life and property as well as threats through ancillary
effects like reduced air quality. The aim of this thesis is to develop
techniques to help combat the impacts of wildfires by improving wildfire
modeling capabilities by using satellite fire observations. Already much work
has been done in this direction by other researchers. Our work seeks to expand
the body of knowledge using mathematically sound methods to utilize information
about wildfires that considers the uncertainties inherent in the satellite
data.
In this thesis we explore methods for using satellite data to help initialize
and steer wildfire simulations. In particular, we develop a method for
constructing the history of a fire, a new technique for assimilating wildfire
data, and a method for modifying the behavior of a modeled fire by inferring
information about the fuels in the fire domain. These goals rely on being able
to estimate the time a fire first arrived at every location in a geographic
region of interest. Because detailed knowledge of real wildfires is typically
unavailable, the basic procedure for developing and testing the methods in this
thesis will be to first work with simulated data so that the estimates produced
can be compared with known solutions. The methods thus developed are then
applied to real-world scenarios. Analysis of these scenarios shows that the
work with constructing the history of fires and data assimilation improves
improves fire modeling capabilities. The research is significant because it
gives us a better understanding of the capabilities and limitations of using
satellite data to inform wildfire models and it points the way towards new
avenues for modeling fire behavior.
|
106,024 | 106,024 |
Testing Feedforward Neural Networks Training Programs
|
Nowadays, we are witnessing an increasing effort to improve the performance
and trustworthiness of Deep Neural Networks (DNNs), with the aim to enable
their adoption in safety critical systems such as self-driving cars. Multiple
testing techniques are proposed to generate test cases that can expose
inconsistencies in the behavior of DNN models. These techniques assume
implicitly that the training program is bug-free and appropriately configured.
However, satisfying this assumption for a novel problem requires significant
engineering work to prepare the data, design the DNN, implement the training
program, and tune the hyperparameters in order to produce the model for which
current automated test data generators search for corner-case behaviors. All
these model training steps can be error-prone. Therefore, it is crucial to
detect and correct errors throughout all the engineering steps of DNN-based
software systems and not only on the resulting DNN model. In this paper, we
gather a catalog of training issues and based on their symptoms and their
effects on the behavior of the training program, we propose practical
verification routines to detect the aforementioned issues, automatically, by
continuously validating that some important properties of the learning dynamics
hold during the training. Then, we design, TheDeepChecker, an end-to-end
property-based debugging approach for DNN training programs. We assess the
effectiveness of TheDeepChecker on synthetic and real-world buggy DL programs
and compare it with Amazon SageMaker Debugger (SMD). Results show that
TheDeepChecker's on-execution validation of DNN-based program's properties
succeeds in revealing several coding bugs and system misconfigurations, early
on and at a low cost. Moreover, TheDeepChecker outperforms the SMD's offline
rules verification on training logs in terms of detection accuracy and DL bugs
coverage.
|
106,025 | 106,025 |
A Reinforcement Learning Approach to Sensing Design in
Resource-Constrained Wireless Networked Control Systems
|
In this paper, we consider a wireless network of smart sensors (agents) that
monitor a dynamical process and send measurements to a base station that
performs global monitoring and decision-making. Smart sensors are equipped with
both sensing and computation, and can either send raw measurements or process
them prior to transmission. Constrained agent resources raise a fundamental
latency-accuracy trade-off. On the one hand, raw measurements are inaccurate
but fast to produce. On the other hand, data processing on resource-constrained
platforms generates accurate measurements at the cost of non-negligible
computation latency. Further, if processed data are also compressed, latency
caused by wireless communication might be higher for raw measurements. Hence,
it is challenging to decide when and where sensors in the network should
transmit raw measurements or leverage time-consuming local processing. To
tackle this design problem, we propose a Reinforcement Learning approach to
learn an efficient policy that dynamically decides when measurements are to be
processed at each sensor. Effectiveness of our proposed approach is validated
through a numerical simulation with case study on smart sensing motivated by
the Internet of Drones.
|
106,026 | 106,026 |
Strategies for Safe Multi-Armed Bandits with Logarithmic Regret and Risk
|
We investigate a natural but surprisingly unstudied approach to the
multi-armed bandit problem under safety risk constraints. Each arm is
associated with an unknown law on safety risks and rewards, and the learner's
goal is to maximise reward whilst not playing unsafe arms, as determined by a
given threshold on the mean risk.
We formulate a pseudo-regret for this setting that enforces this safety
constraint in a per-round way by softly penalising any violation, regardless of
the gain in reward due to the same. This has practical relevance to scenarios
such as clinical trials, where one must maintain safety for each round rather
than in an aggregated sense.
We describe doubly optimistic strategies for this scenario, which maintain
optimistic indices for both safety risk and reward. We show that schema based
on both frequentist and Bayesian indices satisfy tight gap-dependent
logarithmic regret bounds, and further that these play unsafe arms only
logarithmically many times in total. This theoretical analysis is complemented
by simulation studies demonstrating the effectiveness of the proposed schema,
and probing the domains in which their use is appropriate.
|
106,027 | 106,027 |
Identifying Exoplanets with Machine Learning Methods: A Preliminary
Study
|
The discovery of habitable exoplanets has long been a heated topic in
astronomy. Traditional methods for exoplanet identification include the wobble
method, direct imaging, gravitational microlensing, etc., which not only
require a considerable investment of manpower, time, and money, but also are
limited by the performance of astronomical telescopes. In this study, we
proposed the idea of using machine learning methods to identify exoplanets. We
used the Kepler dataset collected by NASA from the Kepler Space Observatory to
conduct supervised learning, which predicts the existence of exoplanet
candidates as a three-categorical classification task, using decision tree,
random forest, na\"ive Bayes, and neural network; we used another NASA dataset
consisted of the confirmed exoplanets data to conduct unsupervised learning,
which divides the confirmed exoplanets into different clusters, using k-means
clustering. As a result, our models achieved accuracies of 99.06%, 92.11%,
88.50%, and 99.79%, respectively, in the supervised learning task and
successfully obtained reasonable clusters in the unsupervised learning task.
|
106,028 | 106,028 |
Analysis of Sparse Subspace Clustering: Experiments and Random
Projection
|
Clustering can be defined as the process of assembling objects into a number
of groups whose elements are similar to each other in some manner. As a
technique that is used in many domains, such as face clustering, plant
categorization, image segmentation, document classification, clustering is
considered one of the most important unsupervised learning problems. Scientists
have surveyed this problem for years and developed different techniques that
can solve it, such as k-means clustering. We analyze one of these techniques: a
powerful clustering algorithm called Sparse Subspace Clustering. We demonstrate
several experiments using this method and then introduce a new approach that
can reduce the computational time required to perform sparse subspace
clustering.
|
106,029 | 106,029 |
SkeleVision: Towards Adversarial Resiliency of Person Tracking with
Multi-Task Learning
|
Person tracking using computer vision techniques has wide ranging
applications such as autonomous driving, home security and sports analytics.
However, the growing threat of adversarial attacks raises serious concerns
regarding the security and reliability of such techniques. In this work, we
study the impact of multi-task learning (MTL) on the adversarial robustness of
the widely used SiamRPN tracker, in the context of person tracking.
Specifically, we investigate the effect of jointly learning with semantically
analogous tasks of person tracking and human keypoint detection. We conduct
extensive experiments with more powerful adversarial attacks that can be
physically realizable, demonstrating the practical value of our approach. Our
empirical study with simulated as well as real-world datasets reveals that
training with MTL consistently makes it harder to attack the SiamRPN tracker,
compared to typically training only on the single task of person tracking.
|
106,030 | 106,030 |
Path Development Network with Finite-dimensional Lie Group
Representation
|
The path signature, a mathematically principled and universal feature of
sequential data, leads to a performance boost of deep learning-based models in
various sequential data tasks as a complimentary feature. However, it suffers
from the curse of dimensionality when the path dimension is high. To tackle
this problem, we propose a novel, trainable path development layer, which
exploits representations of sequential data with the help of finite-dimensional
matrix Lie groups. We also design the backpropagation algorithm of the
development layer via an optimisation method on manifolds known as
trivialisation. Numerical experiments demonstrate that the path development
consistently and significantly outperforms, in terms of accuracy and
dimensionality, signature features on several empirical datasets. Moreover,
stacking the LSTM with the development layer with a suitable matrix Lie group
is empirically proven to alleviate the gradient issues of LSTMs and the
resulting hybrid model achieves the state-of-the-art performance.
|
106,031 | 106,031 |
Modeling Dynamic User Preference via Dictionary Learning for Sequential
Recommendation
|
Capturing the dynamics in user preference is crucial to better predict user
future behaviors because user preferences often drift over time. Many existing
recommendation algorithms -- including both shallow and deep ones -- often
model such dynamics independently, i.e., user static and dynamic preferences
are not modeled under the same latent space, which makes it difficult to fuse
them for recommendation. This paper considers the problem of embedding a user's
sequential behavior into the latent space of user preferences, namely
translating sequence to preference. To this end, we formulate the sequential
recommendation task as a dictionary learning problem, which learns: 1) a shared
dictionary matrix, each row of which represents a partial signal of user
dynamic preferences shared across users; and 2) a posterior distribution
estimator using a deep autoregressive model integrated with Gated Recurrent
Unit (GRU), which can select related rows of the dictionary to represent a
user's dynamic preferences conditioned on his/her past behaviors. Qualitative
studies on the Netflix dataset demonstrate that the proposed method can capture
the user preference drifts over time and quantitative studies on multiple
real-world datasets demonstrate that the proposed method can achieve higher
accuracy compared with state-of-the-art factorization and neural sequential
recommendation methods. The code is available at
https://github.com/cchao0116/S2PNM-TKDE2021.
|
106,032 | 106,032 |
Variational message passing for online polynomial NARMAX identification
|
We propose a variational Bayesian inference procedure for online nonlinear
system identification. For each output observation, a set of parameter
posterior distributions is updated, which is then used to form a posterior
predictive distribution for future outputs. We focus on the class of polynomial
NARMAX models, which we cast into probabilistic form and represent in terms of
a Forney-style factor graph. Inference in this graph is efficiently performed
by a variational message passing algorithm. We show empirically that our
variational Bayesian estimator outperforms an online recursive least-squares
estimator, most notably in small sample size settings and low noise regimes,
and performs on par with an iterative least-squares estimator trained offline.
|
106,033 | 106,033 |
Speaker adaptation for Wav2vec2 based dysarthric ASR
|
Dysarthric speech recognition has posed major challenges due to lack of
training data and heavy mismatch in speaker characteristics. Recent ASR systems
have benefited from readily available pretrained models such as wav2vec2 to
improve the recognition performance. Speaker adaptation using fMLLR and
xvectors have provided major gains for dysarthric speech with very little
adaptation data. However, integration of wav2vec2 with fMLLR features or
xvectors during wav2vec2 finetuning is yet to be explored. In this work, we
propose a simple adaptation network for fine-tuning wav2vec2 using fMLLR
features. The adaptation network is also flexible to handle other speaker
adaptive features such as xvectors. Experimental analysis show steady
improvements using our proposed approach across all impairment severity levels
and attains 57.72\% WER for high severity in UASpeech dataset. We also
performed experiments on German dataset to substantiate the consistency of our
proposed approach across diverse domains.
|
106,034 | 106,034 |
Revealing the real-world CO2 emission reduction of ridesplitting and its
determinants based on machine learning
|
Ridesplitting, which is a form of pooled ridesourcing service, has great
potential to alleviate the negative impacts of ridesourcing on the environment.
However, most existing studies only explored its theoretical environmental
benefits based on optimization models and simulations. To put into practice,
this study aims to reveal the real-world emission reduction of ridesplitting
and its determinants based on the observed data of ridesourcing in Chengdu,
China. Integrating the trip data with the COPERT model, this study calculates
the CO2 emissions of shared rides (ridesplitting) and their substituted single
rides (regular ridesourcing) to estimate the CO2 emission reduction of each
ridesplitting trip. The results show that not all ridesplitting trips reduce
emissions from ridesourcing in the real world. The CO2 emission reduction rate
of ridesplitting varies from trip to trip, averaging at 43.15g/km. Then, the
interpretable machine learning models, gradient boosting machines, are applied
to explore the relationship between the CO2 emission reduction rate of
ridesplitting and its determinants. Based on the SHapley Additive exPlanations
method, the overlap rate and detour rate of shared rides are identified to be
the most important factors that determine the CO2 emission reduction rate of
ridesplitting. Increasing the overlap rate, the number of shared rides, average
speed, and ride distance ratio and decreasing the detour rate, actual trip
distance, ride distance gap can increase the CO2 emission reduction rate of
ridesplitting. In addition, nonlinear effects and interactions of several key
factors are examined through the partial dependence plots. This study provides
a scientific method for the government and ridesourcing companies to better
assess and optimize the environmental benefits of ridesplitting.
|
106,035 | 106,035 |
Distributional Gradient Boosting Machines
|
We present a unified probabilistic gradient boosting framework for regression
tasks that models and predicts the entire conditional distribution of a
univariate response variable as a function of covariates. Our likelihood-based
approach allows us to either model all conditional moments of a parametric
distribution, or to approximate the conditional cumulative distribution
function via Normalizing Flows. As underlying computational backbones, our
framework is based on XGBoost and LightGBM. Modelling and predicting the entire
conditional distribution greatly enhances existing tree-based gradient boosting
implementations, as it allows to create probabilistic forecasts from which
prediction intervals and quantiles of interest can be derived. Empirical
results show that our framework achieves state-of-the-art forecast accuracy.
|
106,036 | 106,036 |
Paoding: Supervised Robustness-preserving Data-free Neural Network
Pruning
|
When deploying pre-trained neural network models in real-world applications,
model consumers often encounter resource-constraint platforms such as mobile
and smart devices. They typically use the pruning technique to reduce the size
and complexity of the model, generating a lighter one with less resource
consumption. Nonetheless, most existing pruning methods are proposed with a
premise that the model after being pruned has a chance to be fine-tuned or even
retrained based on the original training data. This may be unrealistic in
practice, as the data controllers are often reluctant to provide their model
consumers with the original data. In this work, we study the neural network
pruning in the \emph{data-free} context, aiming to yield lightweight models
that are not only accurate in prediction but also robust against undesired
inputs in open-world deployments. Considering the absence of the fine-tuning
and retraining that can fix the mis-pruned units, we replace the traditional
aggressive one-shot strategy with a conservative one that treats the pruning as
a progressive process. We propose a pruning method based on stochastic
optimization that uses robustness-related metrics to guide the pruning process.
Our method is implemented as a Python package named \textsc{Paoding} and
evaluated with a series of experiments on diverse neural network models. The
experimental results show that it significantly outperforms existing one-shot
data-free pruning approaches in terms of robustness preservation and accuracy.
|
106,037 | 106,037 |
HLDC: Hindi Legal Documents Corpus
|
Many populous countries including India are burdened with a considerable
backlog of legal cases. Development of automated systems that could process
legal documents and augment legal practitioners can mitigate this. However,
there is a dearth of high-quality corpora that is needed to develop such
data-driven systems. The problem gets even more pronounced in the case of low
resource languages such as Hindi. In this resource paper, we introduce the
Hindi Legal Documents Corpus (HLDC), a corpus of more than 900K legal documents
in Hindi. Documents are cleaned and structured to enable the development of
downstream applications. Further, as a use-case for the corpus, we introduce
the task of bail prediction. We experiment with a battery of models and propose
a Multi-Task Learning (MTL) based model for the same. MTL models use
summarization as an auxiliary task along with bail prediction as the main task.
Experiments with different models are indicative of the need for further
research in this area. We release the corpus and model implementation code with
this paper: https://github.com/Exploration-Lab/HLDC
|
106,038 | 106,038 |
Efficient comparison of sentence embeddings
|
The domain of natural language processing (NLP), which has greatly evolved
over the last years, has highly benefited from the recent developments in word
and sentence embeddings. Such embeddings enable the transformation of complex
NLP tasks, like semantic similarity or Question and Answering (Q&A), into much
simpler to perform vector comparisons. However, such a problem transformation
raises new challenges like the efficient comparison of embeddings and their
manipulation. In this work, we will discuss about various word and sentence
embeddings algorithms, we will select a sentence embedding algorithm, BERT, as
our algorithm of choice and we will evaluate the performance of two vector
comparison approaches, FAISS and Elasticsearch, in the specific problem of
sentence embeddings. According to the results, FAISS outperforms Elasticsearch
when used in a centralized environment with only one node, especially when big
datasets are included.
|
106,039 | 106,039 |
AdaSmooth: An Adaptive Learning Rate Method based on Effective Ratio
|
It is well known that we need to choose the hyper-parameters in Momentum,
AdaGrad, AdaDelta, and other alternative stochastic optimizers. While in many
cases, the hyper-parameters are tuned tediously based on experience becoming
more of an art than science. We present a novel per-dimension learning rate
method for gradient descent called AdaSmooth. The method is insensitive to
hyper-parameters thus it requires no manual tuning of the hyper-parameters like
Momentum, AdaGrad, and AdaDelta methods. We show promising results compared to
other methods on different convolutional neural networks, multi-layer
perceptron, and alternative machine learning tasks. Empirical results
demonstrate that AdaSmooth works well in practice and compares favorably to
other stochastic optimization methods in neural networks.
|
106,040 | 106,040 |
Intelligence at the Extreme Edge: A Survey on Reformable TinyML
|
The rapid miniaturization of Machine Learning (ML) for low powered processing
has opened gateways to provide cognition at the extreme edge (E.g., sensors and
actuators). Dubbed Tiny Machine Learning (TinyML), this upsurging research
field proposes to democratize the use of Machine Learning (ML) and Deep
Learning (DL) on frugal Microcontroller Units (MCUs). MCUs are highly
energy-efficient pervasive devices capable of operating with less than a few
Milliwatts of power. Nevertheless, many solutions assume that TinyML can only
run inference. Despite this, growing interest in TinyML has led to work that
makes them reformable, i.e., work that permits TinyML to improve once deployed.
In line with this, roadblocks in MCU based solutions in general, such as
reduced physical access and long deployment periods of MCUs, deem reformable
TinyML to play a significant part in more effective solutions. In this work, we
present a survey on reformable TinyML solutions with the proposal of a novel
taxonomy for ease of separation. Here, we also discuss the suitability of each
hierarchical layer in the taxonomy for allowing reformability. In addition to
these, we explore the workflow of TinyML and analyze the identified deployment
schemes and the scarcely available benchmarking tools. Furthermore, we discuss
how reformable TinyML can impact a few selected industrial areas and discuss
the challenges and future directions.
|
106,041 | 106,041 |
Chordal Sparsity for Lipschitz Constant Estimation of Deep Neural
Networks
|
Lipschitz constants of neural networks allow for guarantees of robustness in
image classification, safety in controller design, and generalizability beyond
the training data. As calculating Lipschitz constants is NP-hard, techniques
for estimating Lipschitz constants must navigate the trade-off between
scalability and accuracy. In this work, we significantly push the scalability
frontier of a semidefinite programming technique known as LipSDP while
achieving zero accuracy loss. We first show that LipSDP has chordal sparsity,
which allows us to derive a chordally sparse formulation that we call
Chordal-LipSDP. The key benefit is that the main computational bottleneck of
LipSDP, a large semidefinite constraint, is now decomposed into an equivalent
collection of smaller ones: allowing Chordal-LipSDP to outperform LipSDP
particularly as the network depth grows. Moreover, our formulation uses a
tunable sparsity parameter that enables one to gain tighter estimates without
incurring a significant computational cost. We illustrate the scalability of
our approach through extensive numerical experiments.
|
106,042 | 106,042 |
Production of Categorical Data Verifying Differential Privacy:
Conception and Applications to Machine Learning
|
Private and public organizations regularly collect and analyze digitalized
data about their associates, volunteers, clients, etc. However, because most
personal data are sensitive, there is a key challenge in designing
privacy-preserving systems. To tackle privacy concerns, research communities
have proposed different methods to preserve privacy, with Differential privacy
(DP) standing out as a formal definition that allows quantifying the
privacy-utility trade-off. Besides, with the local DP (LDP) model, users can
sanitize their data locally before transmitting it to the server. The objective
of this thesis is thus two-fold: O$_1$) To improve the utility and privacy in
multiple frequency estimates under LDP guarantees, which is fundamental to
statistical learning. And O$_2$) To assess the privacy-utility trade-off of
machine learning (ML) models trained over differentially private data. For
O$_1$, we first tackled the problem from two "multiple" perspectives, i.e.,
multiple attributes and multiple collections throughout time, while focusing on
utility. Secondly, we focused our attention on the multiple attributes aspect
only, in which we proposed a solution focusing on privacy while preserving
utility. In both cases, we demonstrate through analytical and experimental
validations the advantages of our proposed solutions over state-of-the-art LDP
protocols. For O$_2$, we empirically evaluated ML-based solutions designed to
solve real-world problems while ensuring DP guarantees. Indeed, we mainly used
the input data perturbation setting from the privacy-preserving ML literature.
This is the situation in which the whole dataset is sanitized independently
and, thus, we implemented LDP algorithms from the perspective of the
centralized data owner. In all cases, we concluded that differentially private
ML models achieve nearly the same utility metrics as non-private ones.
|
106,043 | 106,043 |
Adversarial Neon Beam: Robust Physical-World Adversarial Attack to DNNs
|
In the physical world, light affects the performance of deep neural networks.
Nowadays, many products based on deep neural network have been put into daily
life. There are few researches on the effect of light on the performance of
deep neural network models. However, the adversarial perturbations generated by
light may have extremely dangerous effects on these systems. In this work, we
propose an attack method called adversarial neon beam (AdvNB), which can
execute the physical attack by obtaining the physical parameters of adversarial
neon beams with very few queries. Experiments show that our algorithm can
achieve advanced attack effect in both digital test and physical test. In the
digital environment, 99.3% attack success rate was achieved, and in the
physical environment, 100% attack success rate was achieved. Compared with the
most advanced physical attack methods, our method can achieve better physical
perturbation concealment. In addition, by analyzing the experimental data, we
reveal some new phenomena brought about by the adversarial neon beam attack.
|
106,044 | 106,044 |
A Differential Evolution-Enhanced Latent Factor Analysis Model for
High-dimensional and Sparse Data
|
High-dimensional and sparse (HiDS) matrices are frequently adopted to
describe the complex relationships in various big data-related systems and
applications. A Position-transitional Latent Factor Analysis (PLFA) model can
accurately and efficiently represent an HiDS matrix. However, its involved
latent factors are optimized by stochastic gradient descent with the specific
gradient direction step-by-step, which may cause a suboptimal solution. To
address this issue, this paper proposes a Sequential-Group-Differential-
Evolution (SGDE) algorithm to refine the latent factors optimized by a PLFA
model, thereby achieving a highly-accurate SGDE-PLFA model to HiDS matrices. As
demonstrated by the experiments on four HiDS matrices, a SGDE-PLFA model
outperforms the state-of-the-art models.
|
106,045 | 106,045 |
Accurate Online Posterior Alignments for Principled
Lexically-Constrained Decoding
|
Online alignment in machine translation refers to the task of aligning a
target word to a source word when the target sequence has only been partially
decoded. Good online alignments facilitate important applications such as
lexically constrained translation where user-defined dictionaries are used to
inject lexical constraints into the translation model. We propose a novel
posterior alignment technique that is truly online in its execution and
superior in terms of alignment error rates compared to existing methods. Our
proposed inference technique jointly considers alignment and token
probabilities in a principled manner and can be seamlessly integrated within
existing constrained beam-search decoding algorithms. On five language pairs,
including two distant language pairs, we achieve consistent drop in alignment
error rates. When deployed on seven lexically constrained translation tasks, we
achieve significant improvements in BLEU specifically around the constrained
positions.
|
106,046 | 106,046 |
Dimensionless machine learning: Imposing exact units equivariance
|
Units equivariance is the exact symmetry that follows from the requirement
that relationships among measured quantities of physics relevance must obey
self-consistent dimensional scalings. Here, we employ dimensional analysis and
ideas from equivariant machine learning to provide a two stage learning
procedure for units-equivariant machine learning. For a given learning task, we
first construct a dimensionless version of its inputs using classic results
from dimensional analysis, and then perform inference in the dimensionless
space. Our approach can be used to impose units equivariance across a broad
range of machine learning methods which are equivariant to rotations and other
groups. We discuss the in-sample and out-of-sample prediction accuracy gains
one can obtain in contexts like symbolic regression and emulation, where
symmetry is important. We illustrate our approach with simple numerical
examples involving dynamical systems in physics and ecology.
|
106,047 | 106,047 |
Learning List-wise Representation in Reinforcement Learning for Ads
Allocation with Multiple Auxiliary Tasks
|
With the recent prevalence of reinforcement learning (RL), there have been
tremendous interests in utilizing RL for ads allocation in recommendation
platforms (e.g., e-commerce and news feed sites). To achieve better allocation,
the input of recent RL-based ads allocation methods is upgraded from point-wise
single item to list-wise item arrangement. However, this also results in a
high-dimensional space of state-action pairs, making it difficult to learn
list-wise representations with good generalization ability. This further
hinders the exploration of RL agents and causes poor sample efficiency. To
address this problem, we propose a novel RL-based approach for ads allocation
which learns better list-wise representations by leveraging task-specific
signals on Meituan food delivery platform. Specifically, we propose three
different auxiliary tasks based on reconstruction, prediction, and contrastive
learning respectively according to prior domain knowledge on ads allocation. We
conduct extensive experiments on Meituan food delivery platform to evaluate the
effectiveness of the proposed auxiliary tasks. Both offline and online
experimental results show that the proposed method can learn better list-wise
representations and achieve higher revenue for the platform compared to the
state-of-the-art baselines.
|
106,048 | 106,048 |
Class-Incremental Learning by Knowledge Distillation with Adaptive
Feature Consolidation
|
We present a novel class incremental learning approach based on deep neural
networks, which continually learns new tasks with limited memory for storing
examples in the previous tasks. Our algorithm is based on knowledge
distillation and provides a principled way to maintain the representations of
old models while adjusting to new tasks effectively. The proposed method
estimates the relationship between the representation changes and the resulting
loss increases incurred by model updates. It minimizes the upper bound of the
loss increases using the representations, which exploits the estimated
importance of each feature map within a backbone model. Based on the
importance, the model restricts updates of important features for robustness
while allowing changes in less critical features for flexibility. This
optimization strategy effectively alleviates the notorious catastrophic
forgetting problem despite the limited accessibility of data in the previous
tasks. The experimental results show significant accuracy improvement of the
proposed algorithm over the existing methods on the standard datasets. Code is
available.
|
106,049 | 106,049 |
AutoProtoNet: Interpretability for Prototypical Networks
|
In meta-learning approaches, it is difficult for a practitioner to make sense
of what kind of representations the model employs. Without this ability, it can
be difficult to both understand what the model knows as well as to make
meaningful corrections. To address these challenges, we introduce AutoProtoNet,
which builds interpretability into Prototypical Networks by training an
embedding space suitable for reconstructing inputs, while remaining convenient
for few-shot learning. We demonstrate how points in this embedding space can be
visualized and used to understand class representations. We also devise a
prototype refinement method, which allows a human to debug inadequate
classification parameters. We use this debugging technique on a custom
classification task and find that it leads to accuracy improvements on a
validation set consisting of in-the-wild images. We advocate for
interpretability in meta-learning approaches and show that there are
interactive ways for a human to enhance meta-learning algorithms.
|
106,050 | 106,050 |
Exploiting Local and Global Features in Transformer-based Extreme
Multi-label Text Classification
|
Extreme multi-label text classification (XMTC) is the task of tagging each
document with the relevant labels from a very large space of predefined
categories. Recently, large pre-trained Transformer models have made
significant performance improvements in XMTC, which typically use the embedding
of the special CLS token to represent the entire document semantics as a global
feature vector, and match it against candidate labels. However, we argue that
such a global feature vector may not be sufficient to represent different
granularity levels of semantics in the document, and that complementing it with
the local word-level features could bring additional gains. Based on this
insight, we propose an approach that combines both the local and global
features produced by Transformer models to improve the prediction power of the
classifier. Our experiments show that the proposed model either outperforms or
is comparable to the state-of-the-art methods on benchmark datasets.
|
106,051 | 106,051 |
Risk-Aware Control and Optimization for High-Renewable Power Grids
|
The transition of the electrical power grid from fossil fuels to renewable
sources of energy raises fundamental challenges to the market-clearing
algorithms that drive its operations. Indeed, the increased stochasticity in
load and the volatility of renewable energy sources have led to significant
increases in prediction errors, affecting the reliability and efficiency of
existing deterministic optimization models. The RAMC project was initiated to
investigate how to move from this deterministic setting into a risk-aware
framework where uncertainty is quantified explicitly and incorporated in the
market-clearing optimizations. Risk-aware market-clearing raises challenges on
its own, primarily from a computational standpoint. This paper reviews how RAMC
approaches risk-aware market clearing and presents some of its innovations in
uncertainty quantification, optimization, and machine learning. Experimental
results on real networks are presented.
|
106,052 | 106,052 |
Model-Free and Model-Based Policy Evaluation when Causality is Uncertain
|
When decision-makers can directly intervene, policy evaluation algorithms
give valid causal estimates. In off-policy evaluation (OPE), there may exist
unobserved variables that both impact the dynamics and are used by the unknown
behavior policy. These "confounders" will introduce spurious correlations and
naive estimates for a new policy will be biased. We develop worst-case bounds
to assess sensitivity to these unobserved confounders in finite horizons when
confounders are drawn iid each period. We demonstrate that a model-based
approach with robust MDPs gives sharper lower bounds by exploiting domain
knowledge about the dynamics. Finally, we show that when unobserved confounders
are persistent over time, OPE is far more difficult and existing techniques
produce extremely conservative bounds.
|
106,053 | 106,053 |
Long-tailed Extreme Multi-label Text Classification with Generated
Pseudo Label Descriptions
|
Extreme Multi-label Text Classification (XMTC) has been a tough challenge in
machine learning research and applications due to the sheer sizes of the label
spaces and the severe data scarce problem associated with the long tail of rare
labels in highly skewed distributions. This paper addresses the challenge of
tail label prediction by proposing a novel approach, which combines the
effectiveness of a trained bag-of-words (BoW) classifier in generating
informative label descriptions under severe data scarce conditions, and the
power of neural embedding based retrieval models in mapping input documents (as
queries) to relevant label descriptions. The proposed approach achieves
state-of-the-art performance on XMTC benchmark datasets and significantly
outperforms the best methods so far in the tail label prediction. We also
provide a theoretical analysis for relating the BoW and neural models w.r.t.
performance lower bound.
|
106,054 | 106,054 |
Dynamic physical activity recommendation on personalised mobile health
information service: A deep reinforcement learning approach
|
Mobile health (mHealth) information service makes healthcare management
easier for users, who want to increase physical activity and improve health.
However, the differences in activity preference among the individual, adherence
problems, and uncertainty of future health outcomes may reduce the effect of
the mHealth information service. The current health service system usually
provides recommendations based on fixed exercise plans that do not satisfy the
user specific needs. This paper seeks an efficient way to make physical
activity recommendation decisions on physical activity promotion in
personalised mHealth information service by establishing data-driven model. In
this study, we propose a real-time interaction model to select the optimal
exercise plan for the individual considering the time-varying characteristics
in maximising the long-term health utility of the user. We construct a
framework for mHealth information service system comprising a personalised AI
module, which is based on the scientific knowledge about physical activity to
evaluate the individual exercise performance, which may increase the awareness
of the mHealth artificial intelligence system. The proposed deep reinforcement
learning (DRL) methodology combining two classes of approaches to improve the
learning capability for the mHealth information service system. A deep learning
method is introduced to construct the hybrid neural network combing long-short
term memory (LSTM) network and deep neural network (DNN) techniques to infer
the individual exercise behavior from the time series data. A reinforcement
learning method is applied based on the asynchronous advantage actor-critic
algorithm to find the optimal policy through exploration and exploitation.
|
106,055 | 106,055 |
A Dynamic Meta-Learning Model for Time-Sensitive Cold-Start
Recommendations
|
We present a novel dynamic recommendation model that focuses on users who
have interactions in the past but turn relatively inactive recently. Making
effective recommendations to these time-sensitive cold-start users is critical
to maintain the user base of a recommender system. Due to the sparse recent
interactions, it is challenging to capture these users' current preferences
precisely. Solely relying on their historical interactions may also lead to
outdated recommendations misaligned with their recent interests. The proposed
model leverages historical and current user-item interactions and dynamically
factorizes a user's (latent) preference into time-specific and time-evolving
representations that jointly affect user behaviors. These latent factors
further interact with an optimized item embedding to achieve accurate and
timely recommendations. Experiments over real-world data help demonstrate the
effectiveness of the proposed time-sensitive cold-start recommendation model.
|
106,056 | 106,056 |
Kernel Extreme Learning Machine Optimized by the Sparrow Search
Algorithm for Hyperspectral Image Classification
|
To improve the classification performance and generalization ability of the
hyperspectral image classification algorithm, this paper uses Multi-Scale Total
Variation (MSTV) to extract the spectral features, local binary pattern (LBP)
to extract spatial features, and feature superposition to obtain the fused
features of hyperspectral images. A new swarm intelligence optimization method
with high convergence and strong global search capability, the Sparrow Search
Algorithm (SSA), is used to optimize the kernel parameters and regularization
coefficients of the Kernel Extreme Learning Machine (KELM). In summary, a
multiscale fusion feature hyperspectral image classification method (MLS-KELM)
is proposed in this paper. The Indian Pines, Pavia University and Houston 2013
datasets were selected to validate the classification performance of MLS-KELM,
and the method was applied to ZY1-02D hyperspectral data. The experimental
results show that MLS-KELM has better classification performance and
generalization ability compared with other popular classification methods, and
MLS-KELM shows its strong robustness in the small sample case.
|
106,057 | 106,057 |
FedGBF: An efficient vertical federated learning framework via gradient
boosting and bagging
|
Federated learning, conducive to solving data privacy and security problems,
has attracted increasing attention recently. However, the existing federated
boosting model sequentially builds a decision tree model with the weak base
learner, resulting in redundant boosting steps and high interactive
communication costs. In contrast, the federated bagging model saves time by
building multi-decision trees in parallel, but it suffers from performance
loss. With the aim of obtaining an outstanding performance with less time cost,
we propose a novel model in a vertically federated setting termed as Federated
Gradient Boosting Forest (FedGBF). FedGBF simultaneously integrates the
boosting and bagging's preponderance by building the decision trees in parallel
as a base learner for boosting. Subsequent to FedGBF, the problem of
hyperparameters tuning is rising. Then we propose the Dynamic FedGBF, which
dynamically changes each forest's parameters and thus reduces the complexity.
Finally, the experiments based on the benchmark datasets demonstrate the
superiority of our method.
|
106,058 | 106,058 |
Towards Web Phishing Detection Limitations and Mitigation
|
Web phishing remains a serious cyber threat responsible for most data
breaches. Machine Learning (ML)-based anti-phishing detectors are seen as an
effective countermeasure, and are increasingly adopted by web-browsers and
software products. However, with an average of 10K phishing links reported per
hour to platforms such as PhishTank and VirusTotal (VT), the deficiencies of
such ML-based solutions are laid bare. We first explore how phishing sites
bypass ML-based detection with a deep dive into 13K phishing pages targeting
major brands such as Facebook. Results show successful evasion is caused by:
(1) use of benign services to obscure phishing URLs; (2) high similarity
between the HTML structures of phishing and benign pages; (3) hiding the
ultimate phishing content within Javascript and running such scripts only on
the client; (4) looking beyond typical credentials and credit cards for new
content such as IDs and documents; (5) hiding phishing content until after
human interaction. We attribute the root cause to the dependency of ML-based
models on the vertical feature space (webpage content). These solutions rely
only on what phishers present within the page itself. Thus, we propose
Anti-SubtlePhish, a more resilient model based on logistic regression. The key
augmentation is the inclusion of a horizontal feature space, which examines
correlation variables between the final render of suspicious pages against what
trusted services have recorded (e.g., PageRank). To defeat (1) and (2), we
correlate information between WHOIS, PageRank, and page analytics. To combat
(3), (4) and (5), we correlate features after rendering the page. Experiments
with 100K phishing/benign sites show promising accuracy (98.8%). We also
obtained 100% accuracy against 0-day phishing pages that were manually crafted,
comparing well to the 0% recorded by VT vendors over the first four days.
|
106,059 | 106,059 |
Bi-fidelity Modeling of Uncertain and Partially Unknown Systems using
DeepONets
|
Recent advances in modeling large-scale complex physical systems have shifted
research focuses towards data-driven techniques. However, generating datasets
by simulating complex systems can require significant computational resources.
Similarly, acquiring experimental datasets can prove difficult as well. For
these systems, often computationally inexpensive, but in general inaccurate,
models, known as the low-fidelity models, are available. In this paper, we
propose a bi-fidelity modeling approach for complex physical systems, where we
model the discrepancy between the true system's response and low-fidelity
response in the presence of a small training dataset from the true system's
response using a deep operator network (DeepONet), a neural network
architecture suitable for approximating nonlinear operators. We apply the
approach to model systems that have parametric uncertainty and are partially
unknown. Three numerical examples are used to show the efficacy of the proposed
approach to model uncertain and partially unknown complex physical systems.
|
106,060 | 106,060 |
A Computational Analysis of Pitch Drift in Unaccompanied Solo Singing
using DBSCAN Clustering
|
Unaccompanied vocalists usually change the tuning unintentionally and end up
with a higher or lower pitch than the starting point during a long performance.
This phenomenon is called pitch drift, which is dependent on various elements,
such as the skill of the performer, and the length and difficulty of the
performance. In this paper, we propose a computational method for measuring
pitch drift in the course of an unaccompanied vocal performance, using pitch
histogram and DBSCAN clustering.
|
106,061 | 106,061 |
On Efficiently Acquiring Annotations for Multilingual Models
|
When tasked with supporting multiple languages for a given problem, two
approaches have arisen: training a model for each language with the annotation
budget divided equally among them, and training on a high-resource language
followed by zero-shot transfer to the remaining languages. In this work, we
show that the strategy of joint learning across multiple languages using a
single model performs substantially better than the aforementioned
alternatives. We also demonstrate that active learning provides additional,
complementary benefits. We show that this simple approach enables the model to
be data efficient by allowing it to arbitrate its annotation budget to query
languages it is less certain on. We illustrate the effectiveness of our
proposed method on a diverse set of tasks: a classification task with 4
languages, a sequence tagging task with 4 languages and a dependency parsing
task with 5 languages. Our proposed method, whilst simple, substantially
outperforms the other viable alternatives for building a model in a
multilingual setting under constrained budgets.
|
106,062 | 106,062 |
A Differentially Private Framework for Deep Learning with Convexified
Loss Functions
|
Differential privacy (DP) has been applied in deep learning for preserving
privacy of the underlying training sets. Existing DP practice falls into three
categories - objective perturbation, gradient perturbation and output
perturbation. They suffer from three main problems. First, conditions on
objective functions limit objective perturbation in general deep learning
tasks. Second, gradient perturbation does not achieve a satisfactory
privacy-utility trade-off due to over-injected noise in each epoch. Third, high
utility of the output perturbation method is not guaranteed because of the
loose upper bound on the global sensitivity of the trained model parameters as
the noise scale parameter. To address these problems, we analyse a tighter
upper bound on the global sensitivity of the model parameters. Under a
black-box setting, based on this global sensitivity, to control the overall
noise injection, we propose a novel output perturbation framework by injecting
DP noise into a randomly sampled neuron (via the exponential mechanism) at the
output layer of a baseline non-private neural network trained with a
convexified loss function. We empirically compare the privacy-utility
trade-off, measured by accuracy loss to baseline non-private models and the
privacy leakage against black-box membership inference (MI) attacks, between
our framework and the open-source differentially private stochastic gradient
descent (DP-SGD) approaches on six commonly used real-world datasets. The
experimental evaluations show that, when the baseline models have observable
privacy leakage under MI attacks, our framework achieves a better
privacy-utility trade-off than existing DP-SGD implementations, given an
overall privacy budget $\epsilon \leq 1$ for a large number of queries.
|
106,063 | 106,063 |
Understanding the unstable convergence of gradient descent
|
Most existing analyses of (stochastic) gradient descent rely on the condition
that for $L$-smooth costs, the step size is less than $2/L$. However, many
works have observed that in machine learning applications step sizes often do
not fulfill this condition, yet (stochastic) gradient descent still converges,
albeit in an unstable manner. We investigate this unstable convergence
phenomenon from first principles, and discuss key causes behind it. We also
identify its main characteristics, and how they interrelate based on both
theory and experiments, offering a principled view toward understanding the
phenomenon.
|
106,064 | 106,064 |
Learning-Based Approaches for Graph Problems: A Survey
|
Over the years, many graph problems specifically those in NP-complete are
studied by a wide range of researchers. Some famous examples include graph
colouring, travelling salesman problem and subgraph isomorphism. Most of these
problems are typically addressed by exact algorithms, approximate algorithms
and heuristics. There are however some drawback for each of these methods.
Recent studies have employed learning-based frameworks such as machine learning
techniques in solving these problems, given that they are useful in discovering
new patterns in structured data that can be represented using graphs. This
research direction has successfully attracted a considerable amount of
attention. In this survey, we provide a systematic review mainly on classic
graph problems in which learning-based approaches have been proposed in
addressing the problems. We discuss the overview of each framework, and provide
analyses based on the design and performance of the framework. Some potential
research questions are also suggested. Ultimately, this survey gives a clearer
insight and can be used as a stepping stone to the research community in
studying problems in this field.
|
106,065 | 106,065 |
Correlation Functions in Random Fully Connected Neural Networks at
Finite Width
|
This article considers fully connected neural networks with Gaussian random
weights and biases and $L$ hidden layers, each of width proportional to a large
parameter $n$. For polynomially bounded non-linearities we give sharp estimates
in powers of $1/n$ for the joint correlation functions of the network output
and its derivatives. Moreover, we obtain exact layerwise recursions for these
correlation functions and solve a number of special cases for classes of
non-linearities including $\mathrm{ReLU}$ and $\tanh$. We find in both cases
that the depth-to-width ratio $L/n$ plays the role of an effective network
depth, controlling both the scale of fluctuations at individual neurons and the
size of inter-neuron correlations. We use this to study a somewhat simplified
version of the so-called exploding and vanishing gradient problem, proving that
this particular variant occurs if and only if $L/n$ is large. Several of the
key ideas in this article were first developed at a physics level of rigor in a
recent monograph with Roberts and Yaida.
|
106,066 | 106,066 |
Data Cards: Purposeful and Transparent Dataset Documentation for
Responsible AI
|
As research and industry moves towards large-scale models capable of numerous
downstream tasks, the complexity of understanding multi-modal datasets that
give nuance to models rapidly increases. A clear and thorough understanding of
a dataset's origins, development, intent, ethical considerations and evolution
becomes a necessary step for the responsible and informed deployment of models,
especially those in people-facing contexts and high-risk domains. However, the
burden of this understanding often falls on the intelligibility, conciseness,
and comprehensiveness of the documentation. It requires consistency and
comparability across the documentation of all datasets involved, and as such
documentation must be treated as a user-centric product in and of itself. In
this paper, we propose Data Cards for fostering transparent, purposeful and
human-centered documentation of datasets within the practical contexts of
industry and research. Data Cards are structured summaries of essential facts
about various aspects of ML datasets needed by stakeholders across a dataset's
lifecycle for responsible AI development. These summaries provide explanations
of processes and rationales that shape the data and consequently the models,
such as upstream sources, data collection and annotation methods; training and
evaluation methods, intended use; or decisions affecting model performance. We
also present frameworks that ground Data Cards in real-world utility and
human-centricity. Using two case studies, we report on desirable
characteristics that support adoption across domains, organizational
structures, and audience groups. Finally, we present lessons learned from
deploying over 20 Data Cards.
|
106,067 | 106,067 |
Faces: AI Blitz XIII Solutions
|
AI Blitz XIII Faces challenge hosted on www.aicrowd.com platform consisted of
five problems: Sentiment Classification, Age Prediction, Mask Prediction, Face
Recognition, and Face De-Blurring. Our team GLaDOS took second place. Here we
present our solutions and results. Code implementation:
https://github.com/ndrwmlnk/ai-blitz-xiii
|
106,068 | 106,068 |
Breaking the De-Pois Poisoning Defense
|
Attacks on machine learning models have been, since their conception, a very
persistent and evasive issue resembling an endless cat-and-mouse game. One
major variant of such attacks is poisoning attacks which can indirectly
manipulate an ML model. It has been observed over the years that the majority
of proposed effective defense models are only effective when an attacker is not
aware of them being employed. In this paper, we show that the attack-agnostic
De-Pois defense is hardly an exception to that rule. In fact, we demonstrate
its vulnerability to the simplest White-Box and Black-Box attacks by an
attacker that knows the structure of the De-Pois defense model. In essence, the
De-Pois defense relies on a critic model that can be used to detect poisoned
data before passing it to the target model. In our work, we break this
poison-protection layer by replicating the critic model and then performing a
composed gradient-sign attack on both the critic and target models
simultaneously -- allowing us to bypass the critic firewall to poison the
target model.
|
106,069 | 106,069 |
pmuBAGE: The Benchmarking Assortment of Generated PMU Data for Power
System Events -- Part I: Overview and Results
|
We present pmuGE (phasor measurement unit Generator of Events), one of the
first data-driven generative model for power system event data. We have trained
this model on thousands of actual events and created a dataset denoted pmuBAGE
(the Benchmarking Assortment of Generated PMU Events). The dataset consists of
almost 1000 instances of labeled event data to encourage benchmark evaluations
on phasor measurement unit (PMU) data analytics. The dataset is available
online for use by any researcher or practitioner in the field. PMU data are
challenging to obtain, especially those covering event periods. Nevertheless,
power system problems have recently seen phenomenal advancements via
data-driven machine learning solutions - solutions created by researchers who
were fortunate enough to obtain such PMU data. A highly accessible standard
benchmarking dataset would enable a drastic acceleration of the development of
successful machine learning techniques in this field. We propose a novel
learning method based on the Event Participation Decomposition of Power System
Events, which makes it possible to learn a generative model of PMU data during
system anomalies. The model can create highly realistic event data without
compromising the differential privacy of the PMUs used to train it. The dataset
is available online for any researcher to use at the pmuBAGE Github Repository
- https://github.com/NanpengYu/pmuBAGE.
Part I - This is part I of a two part paper. In part I, we describe a high
level overview of pmuBAGE, its creation, and the experiments used to test it.
Part II will discuss the exact models used in its generation in far more
detail.
|
106,070 | 106,070 |
Adversarially robust segmentation models learn perceptually-aligned
gradients
|
The effects of adversarial training on semantic segmentation networks has not
been thoroughly explored. While previous work has shown that
adversarially-trained image classifiers can be used to perform image synthesis,
we have yet to understand how best to leverage an adversarially-trained
segmentation network to do the same. Using a simple optimizer, we demonstrate
that adversarially-trained semantic segmentation networks can be used to
perform image inpainting and generation. Our experiments demonstrate that
adversarially-trained segmentation networks are more robust and indeed exhibit
perceptually-aligned gradients which help in producing plausible image
inpaintings. We seek to place additional weight behind the hypothesis that
adversarially robust models exhibit gradients that are more
perceptually-aligned with human vision. Through image synthesis, we argue that
perceptually-aligned gradients promote a better understanding of a neural
network's learned representations and aid in making neural networks more
interpretable.
|
106,071 | 106,071 |
Fitting an immersed submanifold to data via Sussmann's orbit theorem
|
This paper describes an approach for fitting an immersed submanifold of a
finite-dimensional Euclidean space to random samples. The reconstruction
mapping from the ambient space to the desired submanifold is implemented as a
composition of an encoder that maps each point to a tuple of (positive or
negative) times and a decoder given by a composition of flows along finitely
many vector fields starting from a fixed initial point. The encoder supplies
the times for the flows. The encoder-decoder map is obtained by empirical risk
minimization, and a high-probability bound is given on the excess risk relative
to the minimum expected reconstruction error over a given class of
encoder-decoder maps. The proposed approach makes fundamental use of Sussmann's
orbit theorem, which guarantees that the image of the reconstruction map is
indeed contained in an immersed submanifold.
|
106,072 | 106,072 |
A System for Interactive Examination of Learned Security Policies
|
We present a system for interactive examination of learned security policies.
It allows a user to traverse episodes of Markov decision processes in a
controlled manner and to track the actions triggered by security policies.
Similar to a software debugger, a user can continue or or halt an episode at
any time step and inspect parameters and probability distributions of interest.
The system enables insight into the structure of a given policy and in the
behavior of a policy in edge cases. We demonstrate the system with a network
intrusion use case. We examine the evolution of an IT infrastructure's state
and the actions prescribed by security policies while an attack occurs. The
policies for the demonstration have been obtained through a reinforcement
learning approach that includes a simulation system where policies are
incrementally learned and an emulation system that produces statistics that
drive the simulation runs.
|
106,073 | 106,073 |
Proceedings of TDA: Applications of Topological Data Analysis to Data
Science, Artificial Intelligence, and Machine Learning Workshop at SDM 2022
|
Topological Data Analysis (TDA) is a rigorous framework that borrows
techniques from geometric and algebraic topology, category theory, and
combinatorics in order to study the "shape" of such complex high-dimensional
data. Research in this area has grown significantly over the last several years
bringing a deeply rooted theory to bear on practical applications in areas such
as genomics, natural language processing, medicine, cybersecurity, energy, and
climate change. Within some of these areas, TDA has also been used to augment
AI and ML techniques.
We believe there is further utility to be gained in this space that can be
facilitated by a workshop bringing together experts (both theorists and
practitioners) and non-experts. Currently there is an active community of pure
mathematicians with research interests in developing and exploring the
theoretical and computational aspects of TDA. Applied mathematicians and other
practitioners are also present in community but do not represent a majority.
This speaks to the primary aim of this workshop which is to grow a wider
community of interest in TDA. By fostering meaningful exchanges between these
groups, from across the government, academia, and industry, we hope to create
new synergies that can only come through building a mutual comprehensive
awareness of the problem and solution spaces.
|
106,074 | 106,074 |
Proactive Anomaly Detection for Robot Navigation with Multi-Sensor
Fusion
|
Despite the rapid advancement of navigation algorithms, mobile robots often
produce anomalous behaviors that can lead to navigation failures. The ability
to detect such anomalous behaviors is a key component in modern robots to
achieve high-levels of autonomy. Reactive anomaly detection methods identify
anomalous task executions based on the current robot state and thus lack the
ability to alert the robot before an actual failure occurs. Such an alert delay
is undesirable due to the potential damage to both the robot and the
surrounding objects. We propose a proactive anomaly detection network (PAAD)
for robot navigation in unstructured and uncertain environments. PAAD predicts
the probability of future failure based on the planned motions from the
predictive controller and the current observation from the perception module.
Multi-sensor signals are fused effectively to provide robust anomaly detection
in the presence of sensor occlusion as seen in field environments. Our
experiments on field robot data demonstrates superior failure identification
performance than previous methods, and that our model can capture anomalous
behaviors in real-time while maintaining a low false detection rate in
cluttered fields. Code, dataset, and video are available at
https://github.com/tianchenji/PAAD
|
106,075 | 106,075 |
Byzantine-Robust Federated Linear Bandits
|
In this paper, we study a linear bandit optimization problem in a federated
setting where a large collection of distributed agents collaboratively learn a
common linear bandit model. Standard federated learning algorithms applied to
this setting are vulnerable to Byzantine attacks on even a small fraction of
agents. We propose a novel algorithm with a robust aggregation oracle that
utilizes the geometric median. We prove that our proposed algorithm is robust
to Byzantine attacks on fewer than half of agents and achieves a sublinear
$\tilde{\mathcal{O}}({T^{3/4}})$ regret with $\mathcal{O}(\sqrt{T})$ steps of
communication in $T$ steps. Moreover, we make our algorithm differentially
private via a tree-based mechanism. Finally, if the level of corruption is
known to be small, we show that using the geometric median of mean oracle for
robust aggregation further improves the regret bound.
|
106,076 | 106,076 |
Best-Response Bayesian Reinforcement Learning with Bayes-adaptive POMDPs
for Centaurs
|
Centaurs are half-human, half-AI decision-makers where the AI's goal is to
complement the human. To do so, the AI must be able to recognize the goals and
constraints of the human and have the means to help them. We present a novel
formulation of the interaction between the human and the AI as a sequential
game where the agents are modelled using Bayesian best-response models. We show
that in this case the AI's problem of helping bounded-rational humans make
better decisions reduces to a Bayes-adaptive POMDP. In our simulated
experiments, we consider an instantiation of our framework for humans who are
subjectively optimistic about the AI's future behaviour. Our results show that
when equipped with a model of the human, the AI can infer the human's bounds
and nudge them towards better decisions. We discuss ways in which the machine
can learn to improve upon its own limitations as well with the help of the
human. We identify a novel trade-off for centaurs in partially observable
tasks: for the AI's actions to be acceptable to the human, the machine must
make sure their beliefs are sufficiently aligned, but aligning beliefs might be
costly. We present a preliminary theoretical analysis of this trade-off and its
dependence on task structure.
|
106,077 | 106,077 |
Seemo: A new tool for early design window view satisfaction evaluation
in residential buildings
|
People spend approximately 90% of their lives indoors, and thus arguably, the
indoor space design can significantly influence occupant well-being. Adequate
views to the outside are one of the most cited indoor qualities related to
occupant well-being. However, due to urbanization and densification trends,
designers may have difficulties in providing vistas and views to the outside
with an assortment of content, which can support the needs of their occupants.
To better understand occupant view satisfaction and provide reliable design
feedback to architects, existing view satisfaction data must be expanded to
capture a wider variety of view scenarios and occupants. Most related research
remains challenging in architectural practice due to a lack of easy-to-use
early-design analysis tools. However, early assessment of view can be
advantageous as design decisions in early design, such as building orientation,
plan layout, and facade design, can improve the view quality. This paper,
hence, presents results from a 181 participant view satisfaction survey with
590 window views. The survey data is used to train a tree-regression model to
predict view satisfaction. The prediction performance was compared to an
existing view assessment framework through case studies. The result showed that
the new prediction is more accurate to the surveyed result than the framework.
Further, the prediction performance was generally high for most responses,
verifying the reliability. To facilitate view analysis in early design, this
paper describes integrating the satisfaction prediction model and a ray-casting
tool to compute view parameters in the CAD environment.
|
106,078 | 106,078 |
Why Exposure Bias Matters: An Imitation Learning Perspective of Error
Accumulation in Language Generation
|
Current language generation models suffer from issues such as repetition,
incoherence, and hallucinations. An often-repeated hypothesis is that this
brittleness of generation models is caused by the training and the generation
procedure mismatch, also referred to as exposure bias. In this paper, we verify
this hypothesis by analyzing exposure bias from an imitation learning
perspective. We show that exposure bias leads to an accumulation of errors,
analyze why perplexity fails to capture this accumulation, and empirically show
that this accumulation results in poor generation quality. Source code to
reproduce these experiments is available at
https://github.com/kushalarora/quantifying_exposure_bias
|
106,079 | 106,079 |
Revisiting Sliced Wasserstein on Images: From Vectorization to
Convolution
|
The conventional sliced Wasserstein is defined between two probability
measures that have realizations as vectors. When comparing two probability
measures over images, practitioners first need to vectorize images and then
project them to one-dimensional space by using matrix multiplication between
the sample matrix and the projection matrix. After that, the sliced Wasserstein
is evaluated by averaging the two corresponding one-dimensional projected
probability measures. However, this approach has two limitations. The first
limitation is that the spatial structure of images is not captured efficiently
by the vectorization step; therefore, the later slicing process becomes harder
to gather the discrepancy information. The second limitation is memory
inefficiency since each slicing direction is a vector that has the same
dimension as the images. To address these limitations, we propose novel slicing
methods for sliced Wasserstein between probability measures over images that
are based on the convolution operators. We derive convolution sliced
Wasserstein (CSW) and its variants via incorporating stride, dilation, and
non-linear activation function into the convolution operators. We investigate
the metricity of CSW as well as its sample complexity, its computational
complexity, and its connection to conventional sliced Wasserstein distances.
Finally, we demonstrate the favorable performance of CSW over the conventional
sliced Wasserstein in comparing probability measures over images and in
training deep generative modeling on images.
|
106,080 | 106,080 |
Continuous Variable Quantum MNIST Classifiers
|
In this paper, classical and continuous variable (CV) quantum neural network
hybrid multiclassifiers are presented using the MNIST dataset. The combination
of cutoff dimension and probability measurement method in the CV model allows a
quantum circuit to produce output vectors of size equal to n raised to the
power of n where n represents cutoff dimension and m, the number of qumodes.
They are then translated as one-hot encoded labels, padded with an appropriate
number of zeros. The total of eight different classifiers are built using
2,3,...,8 qumodes, based on the binary classifier architecture proposed in
Continuous variable quantum neural networks. The displacement gate and the Kerr
gate in the CV model allow for the bias addition and nonlinear activation
components of classical neural networks to quantum. The classifiers are
composed of a classical feedforward neural network, a quantum data encoding
circuit, and a CV quantum neural network circuit. On a truncated MNIST dataset
of 600 samples, a 4 qumode hybrid classifier achieves 100% training accuracy.
|
106,081 | 106,081 |
Model-Parallel Fourier Neural Operators as Learned Surrogates for
Large-Scale Parametric PDEs
|
Fourier neural operators (FNOs) are a recently introduced neural network
architecture for learning solution operators of partial differential equations
(PDEs), which have been shown to perform significantly better than comparable
deep learning approaches. Once trained, FNOs can achieve speed-ups of multiple
orders of magnitude over conventional numerical PDE solvers. However, due to
the high dimensionality of their input data and network weights, FNOs have so
far only been applied to two-dimensional or small three-dimensional problems.
To remove this limited problem-size barrier, we propose a model-parallel
version of FNOs based on domain-decomposition of both the input data and
network weights. We demonstrate that our model-parallel FNO is able to predict
time-varying PDE solutions of over 3.2 billion variables on Summit using up to
768 GPUs and show an example of training a distributed FNO on the Azure cloud
for simulating multiphase CO$_2$ dynamics in the Earth's subsurface.
|
106,082 | 106,082 |
Learning Linear Symmetries in Data Using Moment Matching
|
It is common in machine learning and statistics to use symmetries derived
from expert knowledge to simplify problems or improve performance, using
methods like data augmentation or penalties. In this paper we consider the
unsupervised and semi-supervised problems of learning such symmetries in a
distribution directly from data in a model-free fashion. We show that in the
worst case this problem is as difficult as the graph automorphism problem.
However, if we restrict to the case where the covariance matrix has unique
eigenvalues, then the eigenvectors will also be eigenvectors of the symmetry
transformation. If we further restrict to finding orthogonal symmetries, then
the eigenvalues will be either be 1 or -1, and the problem reduces to
determining which eigenvectors are which. We develop and compare theoretically
and empirically the effectiveness of different methods of selecting which
eigenvectors should have eigenvalue -1 in the symmetry transformation, and
discuss how to extend this approach to non-orthogonal cases where we have
labels
|
106,083 | 106,083 |
Capturing positive utilities during the estimation of recursive logit
models: A prism-based approach
|
Although the recursive logit (RL) model has been recently popular and has led
to many applications and extensions, an important numerical issue with respect
to the evaluation of value functions remains unsolved. This issue is
particularly significant for model estimation, during which the parameters are
updated every iteration and may violate the model feasible condition. To solve
this numerical issue, this paper proposes a prism-constrained RL (Prism-RL)
model that implicitly restricts the path set by the prism constraint defined
based upon a state-extended network representation. Providing a set of
numerical experiments, we show that the Prism-RL model succeeds in the stable
estimation regardless of the initial and true parameter values and is able to
capture positive utilities. In the real application to a pedestrian network, we
found the positive effect of street green presence on pedestrians. Moreover,
the Prism-RL model achieved higher goodness of fit than the RL model, implying
that the Prism-RL model can also describe more realistic route choice behavior.
|
106,084 | 106,084 |
MLPro: A System for Hosting Crowdsourced Machine Learning Challenges for
Open-Ended Research Problems
|
The task of developing a machine learning (ML) model for a particular problem
is inherently open-ended, and there is an unbounded set of possible solutions.
Steps of the ML development pipeline, such as feature engineering, loss
function specification, data imputation, and dimensionality reduction, require
the engineer to consider an extensive and often infinite array of
possibilities. Successfully identifying high-performing solutions for an
unfamiliar dataset or problem requires a mix of mathematical prowess and
creativity applied towards inventing and repurposing novel ML methods. Here, we
explore the feasibility of hosting crowdsourced ML challenges to facilitate a
breadth-first exploration of open-ended research problems, thereby expanding
the search space of problem solutions beyond what a typical ML team could
viably investigate. We develop MLPro, a system which combines the notion of
open-ended ML coding problems with the concept of an automatic online code
judging platform. To conduct a pilot evaluation of this paradigm, we
crowdsource several open-ended ML challenges to ML and data science
practitioners. We describe results from two separate challenges. We find that
for sufficiently unconstrained and complex problems, many experts submit
similar solutions, but some experts provide unique solutions which outperform
the "typical" solution class. We suggest that automated expert crowdsourcing
systems such as MLPro have the potential to accelerate ML engineering
creativity.
|
106,085 | 106,085 |
Analysis of Joint Speech-Text Embeddings for Semantic Matching
|
Embeddings play an important role in many recent end-to-end solutions for
language processing problems involving more than one data modality. Although
there has been some effort to understand the properties of single-modality
embedding spaces, particularly that of text, their cross-modal counterparts are
less understood. In this work, we study a joint speech-text embedding space
trained for semantic matching by minimizing the distance between paired
utterance and transcription inputs. This was done through dual encoders in a
teacher-student model setup, with a pretrained language model acting as the
teacher and a transformer-based speech encoder as the student. We extend our
method to incorporate automatic speech recognition through both pretraining and
multitask scenarios and found that both approaches improve semantic matching.
Multiple techniques were utilized to analyze and evaluate cross-modal semantic
alignment of the embeddings: a quantitative retrieval accuracy metric,
zero-shot classification to investigate generalizability, and probing of the
encoders to observe the extent of knowledge transfer from one modality to
another.
|
106,086 | 106,086 |
Differentiable Rendering for Synthetic Aperture Radar Imagery
|
There is rising interest in integrating signal and image processing pipelines
into deep learning training to incorporate more domain knowledge. This can lead
to deep neural networks that are trained more robustly and with limited data,
as well as the capability to solve ill-posed inverse problems. In particular,
there is rising interest in differentiable rendering, which allows explicitly
modeling geometric priors and constraints in the optimization pipeline using
first-order methods such as backpropagation. Existing efforts in differentiable
rendering have focused on imagery from electro-optical sensors, particularly
conventional RGB-imagery. In this work, we propose an approach for
differentiable rendering of Synthetic Aperture Radar (SAR) imagery, which
combines methods from 3D computer graphics with neural rendering. We
demonstrate the approach on the inverse graphics problem of 3D Object
Reconstruction from limited SAR imagery using high-fidelity simulated SAR data.
|
106,087 | 106,087 |
Into-TTS : Intonation Template based Prosody Control System
|
Intonations take an important role in delivering the intention of the
speaker. However, current end-to-end TTS systems often fail to model proper
intonations. To alleviate this problem, we propose a novel, intuitive method to
synthesize speech in different intonations using predefined intonation
templates. Prior to the acoustic model training, speech data are automatically
grouped into intonation templates by k-means clustering, according to their
sentence-final F0 contour. Two proposed modules are added to the end-to-end TTS
framework: intonation classifier and intonation encoder. The intonation
classifier recommends a suitable intonation template to the given text. The
intonation encoder, attached to the text encoder output, synthesizes speech
abiding the requested intonation template. Main contributions of our paper are:
(a) an easy-to-use intonation control system covering a wide range of users;
(b) better performance in wrapping speech in a requested intonation with
improved pitch distance and MOS; and (c) feasibility to future integration
between TTS and NLP, TTS being able to utilize contextual information. Audio
samples are available at https://srtts.github.io/IntoTTS.
|
106,088 | 106,088 |
FedSynth: Gradient Compression via Synthetic Data in Federated Learning
|
Model compression is important in federated learning (FL) with large models
to reduce communication cost. Prior works have been focusing on sparsification
based compression that could desparately affect the global model accuracy. In
this work, we propose a new scheme for upstream communication where instead of
transmitting the model update, each client learns and transmits a light-weight
synthetic dataset such that using it as the training data, the model performs
similarly well on the real training data. The server will recover the local
model update via the synthetic data and apply standard aggregation. We then
provide a new algorithm FedSynth to learn the synthetic data locally.
Empirically, we find our method is comparable/better than random masking
baselines in all three common federated learning benchmark datasets.
|
106,089 | 106,089 |
Empirical Analysis of Lifelog Data using Optimal Feature Selection based
Unsupervised Logistic Regression (OFS-ULR) Model with Spark Streaming
|
Recent advancement in the field of pervasive healthcare monitoring systems
causes the generation of a huge amount of lifelog data in real-time. Chronic
diseases are one of the most serious health challenges in developing and
developed countries. According to WHO, this accounts for 73% of all deaths and
60% of the global burden of diseases. Chronic disease classification models are
now harnessing the potential of lifelog data to explore better healthcare
practices. This paper is to construct an optimal feature selection-based
unsupervised logistic regression model (OFS-ULR) to classify chronic diseases.
Since lifelog data analysis is crucial due to its sensitive nature; thus the
conventional classification models show limited performance. Therefore,
designing new classifiers for the classification of chronic diseases using
lifelog data is the need of the age. The vital part of building a good model
depends on pre-processing of the dataset, identifying important features, and
then training a learning algorithm with suitable hyper parameters for better
performance. The proposed approach improves the performance of existing methods
using a series of steps such as (i) removing redundant or invalid instances,
(ii) making the data labelled using clustering and partitioning the data into
classes, (iii) identifying the suitable subset of features by applying either
some domain knowledge or selection algorithm, (iv) hyper parameter tuning for
models to get best results, and (v) performance evaluation using Spark
streaming environment. For this purpose, two-time series datasets are used in
the experiment to compute the accuracy, recall, precision, and f1-score. The
experimental analysis proves the suitability of the proposed approach as
compared to the conventional classifiers and our newly constructed model
achieved highest accuracy and reduced training complexity among all among all.
|
106,090 | 106,090 |
Explainable Online Lane Change Predictions on a Digital Twin with a
Layer Normalized LSTM and Layer-wise Relevance Propagation
|
Artificial Intelligence and Digital Twins play an integral role in driving
innovation in the domain of intelligent driving. Long short-term memory (LSTM)
is a leading driver in the field of lane change prediction for manoeuvre
anticipation. However, the decision-making process of such models is complex
and non-transparent, hence reducing the trustworthiness of the smart solution.
This work presents an innovative approach and a technical implementation for
explaining lane change predictions of layer normalized LSTMs using Layer-wise
Relevance Propagation (LRP). The core implementation includes consuming live
data from a digital twin on a German highway, live predictions and explanations
of lane changes by extending LRP to layer normalized LSTMs, and an interface
for communicating and explaining the predictions to a human user. We aim to
demonstrate faithful, understandable, and adaptable explanations of lane change
prediction to increase the adoption and trustworthiness of AI systems that
involve humans. Our research also emphases that explainability and
state-of-the-art performance of ML models for manoeuvre anticipation go hand in
hand without negatively affecting predictive effectiveness.
|
106,091 | 106,091 |
GraFN: Semi-Supervised Node Classification on Graph with Few Labels via
Non-Parametric Distribution Assignment
|
Despite the success of Graph Neural Networks (GNNs) on various applications,
GNNs encounter significant performance degradation when the amount of
supervision signals, i.e., number of labeled nodes, is limited, which is
expected as GNNs are trained solely based on the supervision obtained from the
labeled nodes. On the other hand,recent self-supervised learning paradigm aims
to train GNNs by solving pretext tasks that do not require any labeled nodes,
and it has shown to even outperform GNNs trained with few labeled nodes.
However, a major drawback of self-supervised methods is that they fall short of
learning class discriminative node representations since no labeled information
is utilized during training. To this end, we propose a novel semi-supervised
method for graphs, GraFN, that leverages few labeled nodes to ensure nodes that
belong to the same class to be grouped together, thereby achieving the best of
both worlds of semi-supervised and self-supervised methods. Specifically, GraFN
randomly samples support nodes from labeled nodes and anchor nodes from the
entire graph. Then, it minimizes the difference between two predicted class
distributions that are non-parametrically assigned by anchor-supports
similarity from two differently augmented graphs. We experimentally show that
GraFN surpasses both the semi-supervised and self-supervised methods in terms
of node classification on real-world graphs. The source code for GraFN is
available at https://github.com/Junseok0207/GraFN.
|
106,092 | 106,092 |
Algorithms for Bayesian network modeling and reliability inference of
complex multistate systems: Part II-Dependent systems
|
In using the Bayesian network (BN) to construct the complex multistate
system's reliability model as described in Part I, the memory storage
requirements of the node probability table (NPT) will exceed the random access
memory (RAM) of the computer. However, the proposed inference algorithm of Part
I is not suitable for the dependent system. This Part II proposes a novel
method for BN reliability modeling and analysis to apply the compression idea
to the complex multistate dependent system. In this Part II, the dependent
nodes and their parent nodes are equivalent to a block, based on which the
multistate joint probability inference algorithm is proposed to calculate the
joint probability distribution of a block's all nodes. Then, based on the
proposed multistate compression algorithm of Part I, the dependent multistate
inference algorithm is proposed for the complex multistate dependent system.
The use and accuracy of the proposed algorithms are demonstrated in case 1.
Finally, the proposed algorithms are applied to the reliability modeling and
analysis of the satellite attitude control system. The results show that both
Part I and Part II's proposed algorithms make the reliability modeling and
analysis of the complex multistate system feasible.
|
106,093 | 106,093 |
Efficient, Uncertainty-based Moderation of Neural Networks Text
Classifiers
|
To maximize the accuracy and increase the overall acceptance of text
classifiers, we propose a framework for the efficient, in-operation moderation
of classifiers' output. Our framework focuses on use cases in which F1-scores
of modern Neural Networks classifiers (ca.~90%) are still inapplicable in
practice. We suggest a semi-automated approach that uses prediction
uncertainties to pass unconfident, probably incorrect classifications to human
moderators. To minimize the workload, we limit the human moderated data to the
point where the accuracy gains saturate and further human effort does not lead
to substantial improvements. A series of benchmarking experiments based on
three different datasets and three state-of-the-art classifiers show that our
framework can improve the classification F1-scores by 5.1 to 11.2% (up to
approx.~98 to 99%), while reducing the moderation load up to 73.3% compared to
a random moderation.
|
106,094 | 106,094 |
MOSRA: Joint Mean Opinion Score and Room Acoustics Speech Quality
Assessment
|
The acoustic environment can degrade speech quality during communication
(e.g., video call, remote presentation, outside voice recording), and its
impact is often unknown. Objective metrics for speech quality have proven
challenging to develop given the multi-dimensionality of factors that affect
speech quality and the difficulty of collecting labeled data. Hypothesizing the
impact of acoustics on speech quality, this paper presents MOSRA: a
non-intrusive multi-dimensional speech quality metric that can predict room
acoustics parameters (SNR, STI, T60, DRR, and C50) alongside the overall mean
opinion score (MOS) for speech quality. By explicitly optimizing the model to
learn these room acoustics parameters, we can extract more informative features
and improve the generalization for the MOS task when the training data is
limited. Furthermore, we also show that this joint training method enhances the
blind estimation of room acoustics, improving the performance of current
state-of-the-art models. An additional side-effect of this joint prediction is
the improvement in the explainability of the predictions, which is a valuable
feature for many applications.
|
106,095 | 106,095 |
Discretely Indexed Flows
|
In this paper we propose Discretely Indexed flows (DIF) as a new tool for
solving variational estimation problems. Roughly speaking, DIF are built as an
extension of Normalizing Flows (NF), in which the deterministic transport
becomes stochastic, and more precisely discretely indexed. Due to the discrete
nature of the underlying additional latent variable, DIF inherit the good
computational behavior of NF: they benefit from both a tractable density as
well as a straightforward sampling scheme, and can thus be used for the dual
problems of Variational Inference (VI) and of Variational density estimation
(VDE). On the other hand, DIF can also be understood as an extension of mixture
density models, in which the constant mixture weights are replaced by flexible
functions. As a consequence, DIF are better suited for capturing distributions
with discontinuities, sharp edges and fine details, which is a main advantage
of this construction. Finally we propose a methodology for constructiong DIF in
practice, and see that DIF can be sequentially cascaded, and cascaded with NF.
|
106,096 | 106,096 |
Deep learning, stochastic gradient descent and diffusion maps
|
Stochastic gradient descent (SGD) is widely used in deep learning due to its
computational efficiency but a complete understanding of why SGD performs so
well remains a major challenge. It has been observed empirically that most
eigenvalues of the Hessian of the loss functions on the loss landscape of
over-parametrized deep networks are close to zero, while only a small number of
eigenvalues are large. Zero eigenvalues indicate zero diffusion along the
corresponding directions. This indicates that the process of minima selection
mainly happens in the relatively low-dimensional subspace corresponding to top
eigenvalues of the Hessian. Although the parameter space is very
high-dimensional, these findings seems to indicate that the SGD dynamics may
mainly live on a low-dimensional manifold. In this paper we pursue a truly data
driven approach to the problem of getting a potentially deeper understanding of
the high-dimensional parameter surface, and in particular of the landscape
traced out by SGD, by analyzing the data generated through SGD, or any other
optimizer for that matter, in order to possibly discovery (local)
low-dimensional representations of the optimization landscape. As our vehicle
for the exploration we use diffusion maps introduced by R. Coifman and
coauthors.
|
106,097 | 106,097 |
Learning to solve Minimum Cost Multicuts efficiently using Edge-Weighted
Graph Convolutional Neural Networks
|
The minimum cost multicut problem is the NP-hard/APX-hard combinatorial
optimization problem of partitioning a real-valued edge-weighted graph such as
to minimize the total cost of the partition. While graph convolutional neural
networks (GNN) have proven to be promising in the context of combinatorial
optimization, most of them are only tailored to or tested on positive-valued
edge weights, i.e. they do not comply to the nature of the multicut problem. We
therefore adapt various GNN architectures including Graph Convolutional
Networks, Signed Graph Convolutional Networks and Graph Isomorphic Networks to
facilitate the efficient encoding of real-valued edge costs. Moreover, we
employ a reformulation of the multicut ILP constraints to a polynomial program
as loss function that allows to learn feasible multicut solutions in a scalable
way. Thus, we provide the first approach towards end-to-end trainable
multicuts. Our findings support that GNN approaches can produce good solutions
in practice while providing lower computation times and largely improved
scalability compared to LP solvers and optimized heuristics, especially when
considering large instances.
|
106,098 | 106,098 |
Training Fully Connected Neural Networks is $\exists\mathbb{R}$-Complete
|
We consider the algorithmic problem of finding the optimal weights and biases
for a two-layer fully connected neural network to fit a given set of data
points. This problem is known as empirical risk minimization in the machine
learning community. We show that the problem is $\exists\mathbb{R}$-complete.
This complexity class can be defined as the set of algorithmic problems that
are polynomial-time equivalent to finding real roots of a polynomial with
integer coefficients. Our results hold even if the following restrictions are
all added simultaneously.
$\bullet$ There are exactly two output neurons.
$\bullet$ There are exactly two input neurons.
$\bullet$ The data has only 13 different labels.
$\bullet$ The number of hidden neurons is a constant fraction of the number
of data points.
$\bullet$ The target training error is zero.
$\bullet$ The ReLU activation function is used.
This shows that even very simple networks are difficult to train. The result
offers an explanation (though far from a complete understanding) on why only
gradient descent is widely successful in training neural networks in practice.
We generalize a recent result by Abrahamsen, Kleist and Miltzow [NeurIPS 2021].
This result falls into a recent line of research that tries to unveil that a
series of central algorithmic problems from widely different areas of computer
science and mathematics are $\exists\mathbb{R}$-complete: This includes the art
gallery problem [JACM/STOC 2018], geometric packing [FOCS 2020], covering
polygons with convex polygons [FOCS 2021], and continuous constraint
satisfaction problems [FOCS 2021].
|
106,099 | 106,099 |
Synthetic Graph Generation to Benchmark Graph Learning
|
Graph learning algorithms have attained state-of-the-art performance on many
graph analysis tasks such as node classification, link prediction, and
clustering. It has, however, become hard to track the field's burgeoning
progress. One reason is due to the very small number of datasets used in
practice to benchmark the performance of graph learning algorithms. This
shockingly small sample size (~10) allows for only limited scientific insight
into the problem.
In this work, we aim to address this deficiency. We propose to generate
synthetic graphs, and study the behaviour of graph learning algorithms in a
controlled scenario. We develop a fully-featured synthetic graph generator that
allows deep inspection of different models. We argue that synthetic graph
generations allows for thorough investigation of algorithms and provides more
insights than overfitting on three citation datasets. In the case study, we
show how our framework provides insight into unsupervised and supervised graph
neural network models.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.