title
stringlengths 7
239
| abstract
stringlengths 7
2.76k
| cs
int64 0
1
| phy
int64 0
1
| math
int64 0
1
| stat
int64 0
1
| quantitative biology
int64 0
1
| quantitative finance
int64 0
1
|
|---|---|---|---|---|---|---|---|
Accurate Real Time Localization Tracking in A Clinical Environment using Bluetooth Low Energy and Deep Learning | Deep learning has started to revolutionize several different industries, and
the applications of these methods in medicine are now becoming more
commonplace. This study focuses on investigating the feasibility of tracking
patients and clinical staff wearing Bluetooth Low Energy (BLE) tags in a
radiation oncology clinic using artificial neural networks (ANNs) and
convolutional neural networks (CNNs). The performance of these networks was
compared to relative received signal strength indicator (RSSI) thresholding and
triangulation. By utilizing temporal information, a combined CNN+ANN network
was capable of correctly identifying the location of the BLE tag with an
accuracy of 99.9%. It outperformed a CNN model (accuracy = 94%), a thresholding
model employing majority voting (accuracy = 95%), and a triangulation
classifier utilizing majority voting (accuracy = 95%). Future studies will seek
to deploy this affordable real time location system in hospitals to improve
clinical workflow, efficiency, and patient safety.
| 1 | 1 | 0 | 0 | 0 | 0 |
Metropolis Sampling | Monte Carlo (MC) sampling methods are widely applied in Bayesian inference,
system simulation and optimization problems. The Markov Chain Monte Carlo
(MCMC) algorithms are a well-known class of MC methods which generate a Markov
chain with the desired invariant distribution. In this document, we focus on
the Metropolis-Hastings (MH) sampler, which can be considered as the atom of
the MCMC techniques, introducing the basic notions and different properties. We
describe in details all the elements involved in the MH algorithm and the most
relevant variants. Several improvements and recent extensions proposed in the
literature are also briefly discussed, providing a quick but exhaustive
overview of the current Metropolis-based sampling's world.
| 0 | 0 | 0 | 1 | 0 | 0 |
Approximate Structure Construction Using Large Statistical Swarms | In this paper we describe a novel local algorithm for large statistical
swarms using "harmonic attractor dynamics", by means of which a swarm can
construct harmonics of the environment. This in turn allows the swarm to
approximately reconstruct desired structures in the environment. The robots
navigate in a discrete environment, completely free of localization, being able
to communicate with other robots in its own discrete cell only, and being able
to sense or take reliable action within a disk of radius $r$ around itself. We
present the mathematics that underlie such dynamics and present initial results
demonstrating the proposed algorithm.
| 1 | 0 | 0 | 1 | 0 | 0 |
Search for Food of Birds, Fish and Insects | This book chapter introduces to the problem to which extent search strategies
of foraging biological organisms can be identified by statistical data analysis
and mathematical modeling. A famous paradigm in this field is the Levy Flight
Hypothesis: It states that under certain mathematical conditions Levy flights,
which are a key concept in the theory of anomalous stochastic processes,
provide an optimal search strategy. This hypothesis may be understood
biologically as the claim that Levy flights represent an evolutionary adaptive
optimal search strategy for foraging organisms. Another interpretation,
however, is that Levy flights emerge from the interaction between a forager and
a given (scale-free) distribution of food sources. These hypotheses are
discussed controversially in the current literature. We give examples and
counterexamples of experimental data and their analyses supporting and
challenging them.
| 0 | 0 | 0 | 0 | 1 | 0 |
Poverty Prediction with Public Landsat 7 Satellite Imagery and Machine Learning | Obtaining detailed and reliable data about local economic livelihoods in
developing countries is expensive, and data are consequently scarce. Previous
work has shown that it is possible to measure local-level economic livelihoods
using high-resolution satellite imagery. However, such imagery is relatively
expensive to acquire, often not updated frequently, and is mainly available for
recent years. We train CNN models on free and publicly available multispectral
daytime satellite images of the African continent from the Landsat 7 satellite,
which has collected imagery with global coverage for almost two decades. We
show that despite these images' lower resolution, we can achieve accuracies
that exceed previous benchmarks.
| 1 | 0 | 0 | 1 | 0 | 0 |
Implicit Regularization in Matrix Factorization | We study implicit regularization when optimizing an underdetermined quadratic
objective over a matrix $X$ with gradient descent on a factorization of $X$. We
conjecture and provide empirical and theoretical evidence that with small
enough step sizes and initialization close enough to the origin, gradient
descent on a full dimensional factorization converges to the minimum nuclear
norm solution.
| 1 | 0 | 0 | 1 | 0 | 0 |
A general theory of singular values with applications to signal denoising | We study the Pareto frontier for two competing norms $\|\cdot\|_X$ and
$\|\cdot\|_Y$ on a vector space. For a given vector $c$, the pareto frontier
describes the possible values of $(\|a\|_X,\|b\|_Y)$ for a decomposition
$c=a+b$. The singular value decomposition of a matrix is closely related to the
Pareto frontier for the spectral and nuclear norm. We will develop a general
theory that extends the notion of singular values of a matrix to arbitrary
finite dimensional euclidean vector spaces equipped with dual norms. This also
generalizes the diagonal singular value decompositions for tensors introduced
by the author in previous work. We can apply the results to denoising, where
$c$ is a noisy signal, $a$ is a sparse signal and $b$ is noise. Applications
include 1D total variation denoising, 2D total variation Rudin-Osher-Fatemi
image denoising, LASSO, basis pursuit denoising and tensor decompositions.
| 1 | 0 | 1 | 1 | 0 | 0 |
A Hybrid Deep Learning Architecture for Privacy-Preserving Mobile Analytics | Deep Neural Networks are increasingly being used in a variety of machine
learning applications applied to user data on the cloud. However, this approach
introduces a number of privacy and efficiency challenges, as the cloud operator
can perform secondary inferences on the available data. Recently, advances in
edge processing have paved the way for more efficient, and private, data
processing at the source for simple tasks and lighter models, though they
remain a challenge for larger, and more complicated models. In this paper, we
present a hybrid approach for breaking down large, complex deep models for
cooperative, privacy-preserving analytics. We do this by breaking down the
popular deep architectures and fine-tune them in a suitable way. We then
evaluate the privacy benefits of this approach based on the information exposed
to the cloud service. We also assess the local inference cost of different
layers on a modern handset for mobile applications. Our evaluations show that
by using certain kind of fine-tuning and embedding techniques and at a small
processing cost, we can greatly reduce the level of information available to
unintended tasks applied to the data features on the cloud, and hence achieving
the desired tradeoff between privacy and performance.
| 1 | 0 | 0 | 0 | 0 | 0 |
Exploiting Apache Spark platform for CMS computing analytics | The CERN IT provides a set of Hadoop clusters featuring more than 5 PBytes of
raw storage with different open-source, user-level tools available for
analytical purposes. The CMS experiment started collecting a large set of
computing meta-data, e.g. dataset, file access logs, since 2015. These records
represent a valuable, yet scarcely investigated, set of information that needs
to be cleaned, categorized and analyzed. CMS can use this information to
discover useful patterns and enhance the overall efficiency of the distributed
data, improve CPU and site utilization as well as tasks completion time. Here
we present evaluation of Apache Spark platform for CMS needs. We discuss two
main use-cases CMS analytics and ML studies where efficient process billions of
records stored on HDFS plays an important role. We demonstrate that both Scala
and Python (PySpark) APIs can be successfully used to execute extremely I/O
intensive queries and provide valuable data insight from collected meta-data.
| 0 | 1 | 0 | 0 | 0 | 0 |
Scalable Co-Optimization of Morphology and Control in Embodied Machines | Evolution sculpts both the body plans and nervous systems of agents together
over time. In contrast, in AI and robotics, a robot's body plan is usually
designed by hand, and control policies are then optimized for that fixed
design. The task of simultaneously co-optimizing the morphology and controller
of an embodied robot has remained a challenge. In psychology, the theory of
embodied cognition posits that behavior arises from a close coupling between
body plan and sensorimotor control, which suggests why co-optimizing these two
subsystems is so difficult: most evolutionary changes to morphology tend to
adversely impact sensorimotor control, leading to an overall decrease in
behavioral performance. Here, we further examine this hypothesis and
demonstrate a technique for "morphological innovation protection", which
temporarily reduces selection pressure on recently morphologically-changed
individuals, thus enabling evolution some time to "readapt" to the new
morphology with subsequent control policy mutations. We show the potential for
this method to avoid local optima and converge to similar highly fit
morphologies across widely varying initial conditions, while sustaining fitness
improvements further into optimization. While this technique is admittedly only
the first of many steps that must be taken to achieve scalable optimization of
embodied machines, we hope that theoretical insight into the cause of
evolutionary stagnation in current methods will help to enable the automation
of robot design and behavioral training -- while simultaneously providing a
testbed to investigate the theory of embodied cognition.
| 1 | 0 | 0 | 0 | 0 | 0 |
Normalizing the Taylor expansion of non-deterministic λ-terms, via parallel reduction of resource vectors | It has been known since Ehrhard and Regnier's seminal work on the Taylor
expansion of {\lambda}-terms that this operation commutes with normalization:
the expansion of a {\lambda}-term is always normalizable and its normal form is
the expansion of the Böhm tree of the term. We generalize this result to the
non-uniform setting of the algebraic {\lambda}-calculus, i.e.
{\lambda}-calculus extended with linear combinations of terms. This requires us
to tackle two difficulties: foremost is the fact that Ehrhard and Regnier's
techniques rely heavily on the uniform, deterministic nature of the ordinary
{\lambda}-calculus, and thus cannot be adapted; second is the absence of any
satisfactory generic extension of the notion of Böhm tree in presence of
quantitative non-determinism, which is reflected by the fact that the Taylor
expansion of an algebraic {\lambda}-term is not always normalizable. Our
solution is to provide a fine grained study of the dynamics of
{\beta}-reduction under Taylor expansion, by introducing a notion of reduction
on resource vectors, i.e. infinite linear combinations of resource
{\lambda}-terms. The latter form the multilinear fragment of the differential
{\lambda}-calculus, and resource vectors are the target of the Taylor expansion
of {\lambda}-terms. We show the reduction of resource vectors contains the
image of any {\beta}-reduction step, from which we deduce that Taylor expansion
and normalization commute on the nose. We moreover identify a class of
algebraic {\lambda}-terms, encompassing both normalizable algebraic
{\lambda}-terms and arbitrary ordinary {\lambda}-terms: the expansion of these
is always normalizable, which guides the definition of a generalization of
Böhm trees to this setting.
| 1 | 0 | 0 | 0 | 0 | 0 |
Quantifying telescope phase discontinuities external to AO-systems by use of Phase Diversity and Focal Plane Sharpening | We propose and apply two methods to estimate pupil plane phase
discontinuities for two realistic scenarios on VLT and Keck. The methods use
both Phase Diversity and a form of image sharpening. For the case of VLT, we
simulate the `low wind effect' (LWE) which is responsible for focal plane
errors in the SPHERE system in low wind and good seeing conditions. We
successfully estimate the simulated LWE using both methods, and show that they
are complimentary to one another. We also demonstrate that single image Phase
Diversity (also known as Phase Retrieval with diversity) is also capable of
estimating the simulated LWE when using the natural de-focus on the SPHERE/DTTS
imager. We demonstrate that Phase Diversity can estimate the LWE to within 30
nm RMS WFE, which is within the allowable tolerances to achieve a target SPHERE
contrast of 10$^{-6}$. Finally, we simulate 153 nm RMS of piston errors on the
mirror segments of Keck and produce NIRC2 images subject to these effects. We
show that a single, diverse image with 1.5 waves (PV) of focus can be used to
estimate this error to within 29 nm RMS WFE, and a perfect correction of our
estimation would increase the Strehl ratio of a NIRC2 image by 12\%
| 0 | 1 | 0 | 0 | 0 | 0 |
Adaptive Multilevel Monte Carlo Approximation of Distribution Functions | We analyse a multilevel Monte Carlo method for the approximation of
distribution functions of univariate random variables. Since, by assumption,
the target distribution is not known explicitly, approximations have to be
used. We provide an asymptotic analysis of the error and the cost of the
algorithm. Furthermore we construct an adaptive version of the algorithm that
does not require any a priori knowledge on weak or strong convergence rates. We
apply the adaptive algorithm to smooth path-independent and path-dependent
functionals and to stopped exit times of SDEs.
| 0 | 0 | 1 | 1 | 0 | 0 |
Constraints on the pre-impact orbits of Solar System giant impactors | We provide a fast method for computing constraints on impactor pre-impact
orbits, applying this to the late giant impacts in the Solar System. These
constraints can be used to make quick, broad comparisons of different collision
scenarios, identifying some immediately as low-probability events, and
narrowing the parameter space in which to target follow-up studies with
expensive N-body simulations. We benchmark our parameter space predictions,
finding good agreement with existing N-body studies for the Moon. We suggest
that high-velocity impact scenarios in the inner Solar System, including all
currently proposed single impact scenarios for the formation of Mercury, should
be disfavoured. This leaves a multiple hit-and-run scenario as the most
probable currently proposed for the formation of Mercury.
| 0 | 1 | 0 | 0 | 0 | 0 |
Dark matter in the Reticulum II dSph: a radio search | We present a deep radio search in the Reticulum II dwarf spheroidal (dSph)
galaxy performed with the Australia Telescope Compact Array. Observations were
conducted at 16 cm wavelength, with an rms sensitivity of 0.01 mJy/beam, and
with the goal of searching for synchrotron emission induced by annihilation or
decay of weakly interacting massive particles (WIMPs). Data were complemented
with observations on large angular scales taken with the KAT-7 telescope. We
find no evidence for a diffuse emission from the dSph and we derive competitive
bounds on the WIMP properties. In addition, we detect more than 200 new
background radio sources. Among them, we show there are two compelling
candidates for being the radio counterpart of the possible gamma-ray emission
reported by other groups using Fermi-LAT data.
| 0 | 1 | 0 | 0 | 0 | 0 |
Relevant change points in high dimensional time series | This paper investigates the problem of detecting relevant change points in
the mean vector, say $\mu_t =(\mu_{1,t},\ldots ,\mu_{d,t})^T$ of a high
dimensional time series $(Z_t)_{t\in \mathbb{Z}}$.
While the recent literature on testing for change points in this context
considers hypotheses for the equality of the means $\mu_h^{(1)}$ and
$\mu_h^{(2)}$ before and after the change points in the different components,
we are interested in a null hypothesis of the form $$ H_0: |\mu^{(1)}_{h} -
\mu^{(2)}_{h} | \leq \Delta_h ~~~\mbox{ for all } ~~h=1,\ldots ,d $$ where
$\Delta_1, \ldots , \Delta_d$ are given thresholds for which a smaller
difference of the means in the $h$-th component is considered to be
non-relevant.
We propose a new test for this problem based on the maximum of squared and
integrated CUSUM statistics and investigate its properties as the sample size
$n$ and the dimension $d$ both converge to infinity. In particular, using
Gaussian approximations for the maximum of a large number of dependent random
variables, we show that on certain points of the boundary of the null
hypothesis a standardised version of the maximum converges weakly to a Gumbel
distribution.
| 0 | 0 | 1 | 1 | 0 | 0 |
Disentangling in Variational Autoencoders with Natural Clustering | Learning representations that disentangle the underlying factors of
variability in data is an intuitive precursor to AI with human-like reasoning.
Consequently, it has been the object of many efforts of the machine learning
community. This work takes a step further in this direction by addressing the
scenario where generative factors present a multimodal distribution due to the
existence of class distinction in the data. We formulate a lower bound on the
joint distribution of inputs and class labels and present N-VAE, a model which
is capable of separating factors of variation which are exclusive to certain
classes from factors that are shared among classes. This model implements the
natural clustering prior through the use of a class-conditioned latent space
and a shared latent space. We show its usefulness for detecting and
disentangling class-dependent generative factors as well as for generating rich
artificial samples.
| 1 | 0 | 0 | 1 | 0 | 0 |
Calibrated Boosting-Forest | Excellent ranking power along with well calibrated probability estimates are
needed in many classification tasks. In this paper, we introduce a technique,
Calibrated Boosting-Forest that captures both. This novel technique is an
ensemble of gradient boosting machines that can support both continuous and
binary labels. While offering superior ranking power over any individual
regression or classification model, Calibrated Boosting-Forest is able to
preserve well calibrated posterior probabilities. Along with these benefits, we
provide an alternative to the tedious step of tuning gradient boosting
machines. We demonstrate that tuning Calibrated Boosting-Forest can be reduced
to a simple hyper-parameter selection. We further establish that increasing
this hyper-parameter improves the ranking performance under a diminishing
return. We examine the effectiveness of Calibrated Boosting-Forest on
ligand-based virtual screening where both continuous and binary labels are
available and compare the performance of Calibrated Boosting-Forest with
logistic regression, gradient boosting machine and deep learning. Calibrated
Boosting-Forest achieved an approximately 48% improvement compared to a
state-of-art deep learning model. Moreover, it achieved around 95% improvement
on probability quality measurement compared to the best individual gradient
boosting machine. Calibrated Boosting-Forest offers a benchmark demonstration
that in the field of ligand-based virtual screening, deep learning is not the
universally dominant machine learning model and good calibrated probabilities
can better facilitate virtual screening process.
| 1 | 0 | 0 | 1 | 0 | 0 |
Phase Diagram of $α$-RuCl$_3$ in an in-plane Magnetic Field | The low-temperature magnetic phases in the layered honeycomb lattice material
$\alpha$-RuCl$_3$ have been studied as a function of in-plane magnetic field.
In zero field this material orders magnetically below 7 K with so-called zigzag
order within the honeycomb planes. Neutron diffraction data show that a
relatively small applied field of 2 T is sufficient to suppress the population
of the magnetic domain in which the zigzag chains run along the field
direction. We found that the intensity of the magnetic peaks due to zigzag
order is continuously suppressed with increasing field until their
disappearance at $\mu_o$H$_c$=8 T. At still higher fields (above 8 T) the
zigzag order is destroyed, while bulk magnetization and heat capacity
measurements suggest that the material enters a state with gapped magnetic
excitations. We discuss the magnetic phase diagram obtained in our study in the
context of a quantum phase transition.
| 0 | 1 | 0 | 0 | 0 | 0 |
Post hoc inference via joint family-wise error rate control | We introduce a general methodology for post hoc inference in a large-scale
multiple testing framework. The approach is called "user-agnostic" in the sense
that the statistical guarantee on the number of correct rejections holds for
any set of candidate items selected by the user (after having seen the data).
This task is investigated by defining a suitable criterion, named the
joint-family-wise-error rate (JER for short). We propose several procedures for
controlling the JER, with a special focus on incorporating dependencies while
adapting to the unknown quantity of signal (via a step-down approach). We show
that our proposed setting incorporates as particular cases a version of the
higher criticism as well as the closed testing based approach of Goeman and
Solari (2011). Our theoretical statements are supported by numerical
experiments.
| 0 | 0 | 1 | 1 | 0 | 0 |
Jamming Resistant Receivers for Massive MIMO | We design jamming resistant receivers to enhance the robustness of a massive
MIMO uplink channel against jamming. In the pilot phase, we estimate not only
the desired channel, but also the jamming channel by exploiting purposely
unused pilot sequences. The jamming channel estimate is used to construct the
linear receive filter to reduce impact that jamming has on the achievable
rates. The performance of the proposed scheme is analytically and numerically
evaluated. These results show that the proposed scheme greatly improves the
rates, as compared to conventional receivers. Moreover, the proposed schemes
still work well with stronger jamming power.
| 1 | 0 | 0 | 0 | 0 | 0 |
Thickening and sickening the SYK model | We discuss higher dimensional generalizations of the 0+1-dimensional
Sachdev-Ye-Kitaev (SYK) model that has recently become the focus of intensive
interdisciplinary studies by, both, the condensed matter and field-theoretical
communities. Unlike the previous constructions where multiple SYK copies would
be coupled to each other and/or hybridized with itinerant fermions via
spatially short-ranged random hopping processes, we study algebraically varying
long-range (spatially and/or temporally) correlated random couplings in the
general d+1 dimensions. Such pertinent topics as translationally-invariant
strong-coupling solutions, emergent reparametrization symmetry, effective
action for fluctuations, chaotic behavior, and diffusive transport (or a lack
thereof) are all addressed. We find that the most appealing properties of the
original SYK model that suggest the existence of its 1+1-dimensional
holographic gravity dual do not survive the aforementioned generalizations,
thus lending no additional support to the hypothetical broad (including
'non-AdS/non-CFT') holographic correspondence.
| 0 | 1 | 0 | 0 | 0 | 0 |
Hidden chiral symmetries in BDI multichannel Kitaev chains | Realistic implementations of the Kitaev chain require, in general, the
introduction of extra internal degrees of freedom. In the present work, we
discuss the presence of hidden BDI symmetries for free Hamiltonians describing
systems with an arbitrary number of internal degrees of freedom. We generalize
results of a spinfull Kitaev chain to construct a Hamiltonian with $n$ internal
degrees of freedom and obtain the corresponding hidden chiral symmetry. As an
explicit application of this generalized result, we exploit by analytical and
numerical calculations the case of a spinful 2-band Kitaev chain, which can
host up to 4 Majorana bound states. We also observe the appearence of minigap
states, when chiral symmetry is broken.
| 0 | 1 | 0 | 0 | 0 | 0 |
Charge Berezinskii-Kosterlitz-Thouless transition in superconducting NbTiN films | A half-century after the discovery of the superconductor-insulator transition
(SIT), one of the fundamental predictions of the theory, the charge
Berezinskii-Kosterlitz-Thouless (BKT) transition that is expected to occur at
the insulating side of the SIT, has remained unobserved. The charge BKT
transition is a phenomenon dual to the vortex BKT transition, which is at the
heart of the very existence of two-dimensional superconductivity as a
zero-resistance state appearing at finite temperatures. The dual picture points
to the possibility of the existence of a superinsulating state endowed with
zero conductance at finite temperature. Here, we report the observation of the
charge BKT transition on the insulating side of the SIT, identified by the
critical behavior of the resistance. We find that the critical temperature of
the charge BKT transition depends on the magnetic field exhibiting first the
fast growth and then passing through the maximum at fields much less than the
upper critical field. Finally, we ascertain the effects of the finite
electrostatic screening length and its divergence at the magnetic field-tuned
approach to the superconductor-insulator transition.
| 0 | 1 | 0 | 0 | 0 | 0 |
Seed-Driven Geo-Social Data Extraction - Full Version | Geo-social data has been an attractive source for a variety of problems such
as mining mobility patterns, link prediction, location recommendation, and
influence maximization. However, new geo-social data is increasingly
unavailable and suffers several limitations. In this paper, we aim to remedy
the problem of effective data extraction from geo-social data sources. We first
identify and categorize the limitations of extracting geo-social data. In order
to overcome the limitations, we propose a novel seed-driven approach that uses
the points of one source as the seed to feed as queries for the others. We
additionally handle differences between, and dynamics within the sources by
proposing three variants for optimizing search radius. Furthermore, we provide
an optimization based on recursive clustering to minimize the number of
requests and an adaptive procedure to learn the specific data distribution of
each source. Our comprehensive experiments with six popular sources show that
our seed-driven approach yields 14.3 times more data overall, while our
request-optimized algorithm retrieves up to 95% of the data with less than 16%
of the requests. Thus, our proposed seed-driven approach set new standards for
effective and efficient extraction of geo-social data.
| 1 | 0 | 0 | 0 | 0 | 0 |
Bayesian random-effects meta-analysis using the bayesmeta R package | The random-effects or normal-normal hierarchical model is commonly utilized
in a wide range of meta-analysis applications. A Bayesian approach to inference
is very attractive in this context, especially when a meta-analysis is based
only on few studies. The bayesmeta R package provides readily accessible tools
to perform Bayesian meta-analyses and generate plots and summaries, without
having to worry about computational details. It allows for flexible prior
specification and instant access to the resulting posterior distributions,
including prediction and shrinkage estimation, and facilitating for example
quick sensitivity checks. The present paper introduces the underlying theory
and showcases its usage.
| 0 | 0 | 0 | 1 | 0 | 0 |
Dirichlet Mixture Model based VQ Performance Prediction for Line Spectral Frequency | In this paper, we continue our previous work on the Dirichlet mixture model
(DMM)-based VQ to derive the performance bound of the LSF VQ. The LSF
parameters are transformed into the $\Delta$LSF domain and the underlying
distribution of the $\Delta$LSF parameters are modelled by a DMM with finite
number of mixture components. The quantization distortion, in terms of the mean
squared error (MSE), is calculated with the high rate theory. The mapping
relation between the perceptually motivated log spectral distortion (LSD) and
the MSE is empirically approximated by a polynomial. With this mapping
function, the minimum required bit rate for transparent coding of the LSF is
estimated.
| 1 | 0 | 0 | 1 | 0 | 0 |
Query K-means Clustering and the Double Dixie Cup Problem | We consider the problem of approximate $K$-means clustering with outliers and
side information provided by same-cluster queries and possibly noisy answers.
Our solution shows that, under some mild assumptions on the smallest cluster
size, one can obtain an $(1+\epsilon)$-approximation for the optimal potential
with probability at least $1-\delta$, where $\epsilon>0$ and $\delta\in(0,1)$,
using an expected number of $O(\frac{K^3}{\epsilon \delta})$ noiseless
same-cluster queries and comparison-based clustering of complexity $O(ndK +
\frac{K^3}{\epsilon \delta})$, here, $n$ denotes the number of points and $d$
the dimension of space. Compared to a handful of other known approaches that
perform importance sampling to account for small cluster sizes, the proposed
query technique reduces the number of queries by a factor of roughly
$O(\frac{K^6}{\epsilon^3})$, at the cost of possibly missing very small
clusters. We extend this settings to the case where some queries to the oracle
produce erroneous information, and where certain points, termed outliers, do
not belong to any clusters. Our proof techniques differ from previous methods
used for $K$-means clustering analysis, as they rely on estimating the sizes of
the clusters and the number of points needed for accurate centroid estimation
and subsequent nontrivial generalizations of the double Dixie cup problem. We
illustrate the performance of the proposed algorithm both on synthetic and real
datasets, including MNIST and CIFAR $10$.
| 0 | 0 | 0 | 1 | 0 | 0 |
On the spectrum of directed uniform and non-uniform hypergraphs | Here, we suggest a method to represent general directed uniform and
non-uniform hypergraphs by different connectivity tensors. We show many results
on spectral properties of undirected hypergraphs also hold for general directed
uniform hypergraphs. Our representation of a connectivity tensor will be very
useful for the further development in spectral theory of directed hypergraphs.
At the end, we have also introduced the concept of weak* irreducible
hypermatrix to better explain connectivity of a directed hypergraph.
| 0 | 0 | 1 | 0 | 0 | 0 |
Inference for partial correlation when data are missing not at random | We introduce uncertainty regions to perform inference on partial correlations
when data are missing not at random. These uncertainty regions are shown to
have a desired asymptotic coverage. Their finite sample performance is
illustrated via simulations and real data example.
| 0 | 0 | 1 | 1 | 0 | 0 |
Interstitial Content Detection | Interstitial content is online content which grays out, or otherwise obscures
the main page content. In this technical report, we discuss exploratory
research into detecting the presence of interstitial content in web pages. We
discuss the use of computer vision techniques to detect interstitials, and the
potential use of these techniques to provide a labelled dataset for machine
learning.
| 1 | 0 | 0 | 0 | 0 | 0 |
Perfect spike detection via time reversal | Spiking neuronal networks are usually simulated with three main simulation
schemes: the classical time-driven and event-driven schemes, and the more
recent hybrid scheme. All three schemes evolve the state of a neuron through a
series of checkpoints: equally spaced in the first scheme and determined
neuron-wise by spike events in the latter two. The time-driven and the hybrid
scheme determine whether the membrane potential of a neuron crosses a threshold
at the end of of the time interval between consecutive checkpoints. Threshold
crossing can, however, occur within the interval even if this test is negative.
Spikes can therefore be missed. The present work derives, implements, and
benchmarks a method for perfect retrospective spike detection. This method can
be applied to neuron models with affine or linear subthreshold dynamics. The
idea behind the method is to propagate the threshold with a time-inverted
dynamics, testing whether the threshold crosses the neuron state to be evolved,
rather than vice versa. Algebraically this translates into a set of
inequalities necessary and sufficient for threshold crossing. This test is
slower than the imperfect one, but faster than an alternative perfect tests
based on bisection or root-finding methods. Comparison confirms earlier results
that the imperfect test rarely misses spikes (less than a fraction $1/10^8$ of
missed spikes) in biologically relevant settings. This study offers an
alternative geometric point of view on neuronal dynamics.
| 0 | 1 | 1 | 0 | 0 | 0 |
A response to: "NIST experts urge caution in use of courtroom evidence presentation method" | A press release from the National Institute of Standards and Technology
(NIST)could potentially impede progress toward improving the analysis of
forensic evidence and the presentation of forensic analysis results in courts
in the United States and around the world. "NIST experts urge caution in use of
courtroom evidence presentation method" was released on October 12, 2017, and
was picked up by the phys.org news service. It argues that, except in
exceptional cases, the results of forensic analyses should not be reported as
"likelihood ratios". The press release, and the journal article by NIST
researchers Steven P. Lund & Harri Iyer on which it is based, identifies some
legitimate points of concern, but makes a strawman argument and reaches an
unjustified conclusion that throws the baby out with the bathwater.
| 0 | 0 | 0 | 1 | 0 | 0 |
Safe Model-based Reinforcement Learning with Stability Guarantees | Reinforcement learning is a powerful paradigm for learning optimal policies
from experimental data. However, to find optimal policies, most reinforcement
learning algorithms explore all possible actions, which may be harmful for
real-world systems. As a consequence, learning algorithms are rarely applied on
safety-critical systems in the real world. In this paper, we present a learning
algorithm that explicitly considers safety, defined in terms of stability
guarantees. Specifically, we extend control-theoretic results on Lyapunov
stability verification and show how to use statistical models of the dynamics
to obtain high-performance control policies with provable stability
certificates. Moreover, under additional regularity assumptions in terms of a
Gaussian process prior, we prove that one can effectively and safely collect
data in order to learn about the dynamics and thus both improve control
performance and expand the safe region of the state space. In our experiments,
we show how the resulting algorithm can safely optimize a neural network policy
on a simulated inverted pendulum, without the pendulum ever falling down.
| 1 | 0 | 0 | 1 | 0 | 0 |
Refining Trace Abstraction using Abstract Interpretation | The CEGAR loop in software model checking notoriously diverges when the
abstraction refinement procedure does not derive a loop invariant. An
abstraction refinement procedure based on an SMT solver is applied to a trace,
i.e., a restricted form of a program (without loops). In this paper, we present
a new abstraction refinement procedure that aims at circumventing this
restriction whenever possible. We apply abstract interpretation to a program
that we derive from the given trace. If the program contains a loop, we are
guaranteed to obtain a loop invariant. We call an SMT solver only in the case
where the abstract interpretation returns an indefinite answer. That is, the
idea is to use abstract interpretation and an SMT solver in tandem. An
experimental evaluation in the setting of trace abstraction indicates the
practical potential of this idea.
| 1 | 0 | 0 | 0 | 0 | 0 |
A partial inverse problem for the Sturm-Liouville operator on the graph with a loop | The Sturm-Liouville operator with singular potentials on the lasso graph is
considered. We suppose that the potential is known a priori on the boundary
edge, and recover the potential on the loop from a part of the spectrum and
some additional data. We prove the uniqueness theorem and provide a
constructive algorithm for the solution of this partial inverse problem.
| 0 | 0 | 1 | 0 | 0 | 0 |
BPjs --- a framework for modeling reactive systems using a scripting language and BP | We describe some progress towards a new common framework for model driven
engineering, based on behavioral programming. The tool we have developed
unifies almost all of the work done in behavioral programming so far, under a
common set of interfaces. Its architecture supports pluggable event selection
strategies, which can make models more intuitive and compact. Program state
space can be traversed using various algorithms, such as DFS and A*.
Furthermore, program state is represented in a way that enables scanning a
state space using parallel and distributed algorithms. Executable models
created with this tool can be directly embedded in Java applications, enabling
a model-first approach to system engineering, where initially a model is
created and verified, and then a working application is gradually built around
the model. The model itself consists of a collection of small scripts written
in JavaScript (hence "BPjs"). Using a variety of case-studies, this paper shows
how the combination of a lenient programming language with formal model
analysis tools creates an efficient way of developing robust complex systems.
Additionally, as we learned from an experimental course we ran, the usage of
JavaScript make practitioners more amenable to using this system and, thus,
model checking and model driven engineering. In addition to providing
infrastructure for development and case-studies in behavioral programming, the
tool is designed to serve as a common platform for research and innovation in
behavioral programming and in model driven engineering in general.
| 1 | 0 | 0 | 0 | 0 | 0 |
Design of Quantum Circuits for Galois Field Squaring and Exponentiation | This work presents an algorithm to generate depth, quantum gate and qubit
optimized circuits for $GF(2^m)$ squaring in the polynomial basis. Further, to
the best of our knowledge the proposed quantum squaring circuit algorithm is
the only work that considers depth as a metric to be optimized. We compared
circuits generated by our proposed algorithm against the state of the art and
determine that they require $50 \%$ fewer qubits and offer gates savings that
range from $37 \%$ to $68 \%$. Further, existing quantum exponentiation are
based on either modular or integer arithmetic. However, Galois arithmetic is a
useful tool to design resource efficient quantum exponentiation circuit
applicable in quantum cryptanalysis. Therefore, we present the quantum circuit
implementation of Galois field exponentiation based on the proposed quantum
Galois field squaring circuit. We calculated a qubit savings ranging between
$44\%$ to $50\%$ and quantum gate savings ranging between $37 \%$ to $68 \%$
compared to identical quantum exponentiation circuit based on existing squaring
circuits.
| 1 | 0 | 0 | 0 | 0 | 0 |
Nearest-Neighbor Sample Compression: Efficiency, Consistency, Infinite Dimensions | We examine the Bayes-consistency of a recently proposed
1-nearest-neighbor-based multiclass learning algorithm. This algorithm is
derived from sample compression bounds and enjoys the statistical advantages of
tight, fully empirical generalization bounds, as well as the algorithmic
advantages of a faster runtime and memory savings. We prove that this algorithm
is strongly Bayes-consistent in metric spaces with finite doubling dimension
--- the first consistency result for an efficient nearest-neighbor sample
compression scheme. Rather surprisingly, we discover that this algorithm
continues to be Bayes-consistent even in a certain infinite-dimensional
setting, in which the basic measure-theoretic conditions on which classic
consistency proofs hinge are violated. This is all the more surprising, since
it is known that $k$-NN is not Bayes-consistent in this setting. We pose
several challenging open problems for future research.
| 1 | 0 | 1 | 1 | 0 | 0 |
AdaGrad stepsizes: Sharp convergence over nonconvex landscapes, from any initialization | Adaptive gradient methods such as AdaGrad and its variants update the
stepsize in stochastic gradient descent on the fly according to the gradients
received along the way; such methods have gained widespread use in large-scale
optimization for their ability to converge robustly, without the need to fine
tune parameters such as the stepsize schedule. Yet, the theoretical guarantees
to date for AdaGrad are for online and convex optimization, which is quite
different from the offline and nonconvex setting where adaptive gradient
methods shine in practice. We bridge this gap by providing strong theoretical
guarantees in batch and stochastic setting, for the convergence of AdaGrad over
smooth, nonconvex landscapes, from any initialization of the stepsize, without
knowledge of Lipschitz constant of the gradient. We show in the stochastic
setting that AdaGrad converges to a stationary point at the optimal
$O(1/\sqrt{N})$ rate (up to a $\log(N)$ factor), and in the batch setting, at
the optimal $O(1/N)$ rate. Moreover, in both settings, the constant in the rate
matches the constant obtained as if the variance of the gradient noise and
Lipschitz constant of the gradient were known in advance and used to tune the
stepsize, up to a logarithmic factor of the mismatch between the optimal
stepsize and the stepsize used to initialize AdaGrad. In particular, our
results imply that AdaGrad is robust to both the unknown Lipschitz constant and
level of stochastic noise on the gradient, in a near-optimal sense. When there
is noise, AdaGrad converges at the rate of $O(1/\sqrt{N})$ with well-tuned
stepsize, and when there is not noise, the same algorithm converges at the rate
of $O(1/N)$ like well-tuned batch gradient descent.
| 0 | 0 | 0 | 1 | 0 | 0 |
On polynomially integrable convex bodies | An infinitely smooth convex body in $\mathbb R^n$ is called polynomially
integrable of degree $N$ if its parallel section functions are polynomials of
degree $N$. We prove that the only smooth convex bodies with this property in
odd dimensions are ellipsoids, if $N\ge n-1$. This is in contrast with the case
of even dimensions and the case of odd dimensions with $N<n-1$, where such
bodies do not exist, as it was recently shown by Agranovsky.
| 0 | 0 | 1 | 0 | 0 | 0 |
Percentile Policies for Tracking of Markovian Random Processes with Asymmetric Cost and Observation | Motivated by wide-ranging applications such as video delivery over networks
using Multiple Description Codes, congestion control, and inventory management,
we study the state-tracking of a Markovian random process with a known
transition matrix and a finite ordered state set. The decision-maker must
select a state as an action at each time step to minimize the total expected
cost. The decision-maker is faced with asymmetries both in cost and
observation: in case the selected state is less than the actual state of the
Markovian process, an under-utilization cost occurs and only partial
observation about the actual state is revealed; otherwise, the decision incurs
an over-utilization cost and reveals full information about the actual state.
We can formulate this problem as a Partially Observable Markov Decision Process
which can be expressed as a dynamic program based on the last full observed
state and the time of full observation. This formulation determines the
sequence of actions to be taken between any two consecutive full observations
of the actual state. However, this DP grows exponentially in the number of
states, with little hope for a computationally feasible solution. We present an
interesting class of computationally tractable policies with a percentile
structure. A generalization of binary search, this class of policies attempt at
any given time to reduce the uncertainty by a given percentage. Among all
percentile policies, we search for the one with the minimum expected cost. The
result of this search is a heuristic policy which we evaluate through numerical
simulations. We show that it outperforms the myopic policies and under some
conditions performs close to the optimal policies. Furthermore, we derive a
lower bound on the cost of the optimal policy which can be computed with low
complexity and give a measure for how close our heuristic policy is to the
optimal policy.
| 1 | 0 | 0 | 0 | 0 | 0 |
On Information Transfer Based Characterization of Power System Stability | In this paper, we present a novel approach to identify the generators and
states responsible for the small-signal stability of power networks. To this
end, the newly developed notion of information transfer between the states of a
dynamical system is used. In particular, using the concept of information
transfer, which characterizes influence between the various states and a linear
combination of states of a dynamical system, we identify the generators and
states which are responsible for causing instability of the power network.
While characterizing influence from state to state, information transfer can
also describe influence from state to modes thereby generalizing the well-known
notion of participation factor while at the same time overcoming some of the
limitations of the participation factor. The developed framework is applied to
study the three bus system identifying various cause of instabilities in the
system. The simulation study is extended to IEEE 39 bus system.
| 1 | 0 | 0 | 0 | 0 | 0 |
Using Session Types for Reasoning About Boundedness in the Pi-Calculus | The classes of depth-bounded and name-bounded processes are fragments of the
pi-calculus for which some of the decision problems that are undecidable for
the full calculus become decidable. P is depth-bounded at level k if every
reduction sequence for P contains successor processes with at most k active
nested restrictions. P is name-bounded at level k if every reduction sequence
for P contains successor processes with at most k active bound names.
Membership of these classes of processes is undecidable. In this paper we use
binary session types to decise two type systems that give a sound
characterization of the properties: If a process is well-typed in our first
system, it is depth-bounded. If a process is well-typed in our second, more
restrictive type system, it will also be name-bounded.
| 1 | 0 | 0 | 0 | 0 | 0 |
Combinatorial Secretary Problems with Ordinal Information | The secretary problem is a classic model for online decision making.
Recently, combinatorial extensions such as matroid or matching secretary
problems have become an important tool to study algorithmic problems in dynamic
markets. Here the decision maker must know the numerical value of each arriving
element, which can be a demanding informational assumption. In this paper, we
initiate the study of combinatorial secretary problems with ordinal
information, in which the decision maker only needs to be aware of a preference
order consistent with the values of arrived elements. The goal is to design
online algorithms with small competitive ratios.
For a variety of combinatorial problems, such as bipartite matching, general
packing LPs, and independent set with bounded local independence number, we
design new algorithms that obtain constant competitive ratios.
For the matroid secretary problem, we observe that many existing algorithms
for special matroid structures maintain their competitive ratios even in the
ordinal model. In these cases, the restriction to ordinal information does not
represent any additional obstacle. Moreover, we show that ordinal variants of
the submodular matroid secretary problems can be solved using algorithms for
the linear versions by extending [Feldman and Zenklusen, 2015]. In contrast, we
provide a lower bound of $\Omega(\sqrt{n}/(\log n))$ for algorithms that are
oblivious to the matroid structure, where $n$ is the total number of elements.
This contrasts an upper bound of $O(\log n)$ in the cardinal model, and it
shows that the technique of thresholding is not sufficient for good algorithms
in the ordinal model.
| 1 | 0 | 0 | 0 | 0 | 0 |
Modeling the Formation of Social Conventions in Multi-Agent Populations | In order to understand the formation of social conventions we need to know
the specific role of control and learning in multi-agent systems. To advance in
this direction, we propose, within the framework of the Distributed Adaptive
Control (DAC) theory, a novel Control-based Reinforcement Learning architecture
(CRL) that can account for the acquisition of social conventions in multi-agent
populations that are solving a benchmark social decision-making problem. Our
new CRL architecture, as a concrete realization of DAC multi-agent theory,
implements a low-level sensorimotor control loop handling the agent's reactive
behaviors (pre-wired reflexes), along with a layer based on model-free
reinforcement learning that maximizes long-term reward. We apply CRL in a
multi-agent game-theoretic task in which coordination must be achieved in order
to find an optimal solution. We show that our CRL architecture is able to both
find optimal solutions in discrete and continuous time and reproduce human
experimental data on standard game-theoretic metrics such as efficiency in
acquiring rewards, fairness in reward distribution and stability of convention
formation.
| 0 | 0 | 0 | 1 | 1 | 0 |
A class of C*-algebraic locally compact quantum groupoids Part I. Motivation and definition | In this series of papers, we develop the theory of a class of locally compact
quantum groupoids, which is motivated by the purely algebraic notion of weak
multiplier Hopf algebras. In this Part I, we provide motivation and formulate
the definition in the C*-algebra framework. Existence of a certain canonical
idempotent element is required and it plays a fundamental role, including the
establishment of the coassociativity of the comultiplication. This class
contains locally compact quantum groups as a subclass.
| 0 | 0 | 1 | 0 | 0 | 0 |
Symmetry-enforced quantum spin Hall insulators in $π$-flux models | We prove a Lieb-Schultz-Mattis theorem for the quantum spin Hall effect
(QSHE) in two-dimensional $\pi$-flux models. In the presence of time reversal,
$U(1)$ charge conservation and magnetic translation (with $\pi$-flux per unit
cell) symmetries, if a generic interacting Hamiltonian has a unique gapped
symmetric ground state at half filling (i.e. an odd number of electrons per
unit cell), it can only be a QSH insulator. In other words, a trivial Mott
insulator is forbidden by symmetries at half filling. We further show that such
a symmetry-enforced QSHE can be realized in cold atoms, by shaking an optical
lattice and applying a time-dependent Zeeman field.
| 0 | 1 | 0 | 0 | 0 | 0 |
Stabilized microwave-frequency transfer using optical phase sensing and actuation | We present a stabilized microwave-frequency transfer technique that is based
on optical phase-sensing and optical phase-actuation. This technique shares
several attributes with optical-frequency transfer and therefore exhibits
several advantages over other microwave-frequency transfer techniques. We
demonstrated stabilized transfer of an 8,000 MHz microwave-frequency signal
over a 166 km metropolitan optical fiber network, achieving a fractional
frequency stability of 6.8x10^-14 Hz/Hz at 1 s integration, and 5.0x10^-16
Hz/Hz at 1.6x10^4 s. This technique is being considered for use on the Square
Kilometre Array SKA1-mid radio telescope.
| 0 | 1 | 0 | 0 | 0 | 0 |
Spectral sets for numerical range | We define and study a numerical-range analogue of the notion of spectral set.
Among the results obtained are a positivity criterion and a dilation theorem,
analogous to those already known for spectral sets. An important difference
from the classical definition is the role played in the new definition by the
base point. We present some examples to illustrate this aspect.
| 0 | 0 | 1 | 0 | 0 | 0 |
Learning of Gaussian Processes in Distributed and Communication Limited Systems | It is of fundamental importance to find algorithms obtaining optimal
performance for learning of statistical models in distributed and communication
limited systems. Aiming at characterizing the optimal strategies, we consider
learning of Gaussian Processes (GPs) in distributed systems as a pivotal
example. We first address a very basic problem: how many bits are required to
estimate the inner-products of Gaussian vectors across distributed machines?
Using information theoretic bounds, we obtain an optimal solution for the
problem which is based on vector quantization. Two suboptimal and more
practical schemes are also presented as substitute for the vector quantization
scheme. In particular, it is shown that the performance of one of the practical
schemes which is called per-symbol quantization is very close to the optimal
one. Schemes provided for the inner-product calculations are incorporated into
our proposed distributed learning methods for GPs. Experimental results show
that with spending few bits per symbol in our communication scheme, our
proposed methods outperform previous zero rate distributed GP learning schemes
such as Bayesian Committee Model (BCM) and Product of experts (PoE).
| 1 | 0 | 0 | 1 | 0 | 0 |
Deep Learning: A Critical Appraisal | Although deep learning has historical roots going back decades, neither the
term "deep learning" nor the approach was popular just over five years ago,
when the field was reignited by papers such as Krizhevsky, Sutskever and
Hinton's now classic (2012) deep network model of Imagenet. What has the field
discovered in the five subsequent years? Against a background of considerable
progress in areas such as speech recognition, image recognition, and game
playing, and considerable enthusiasm in the popular press, I present ten
concerns for deep learning, and suggest that deep learning must be supplemented
by other techniques if we are to reach artificial general intelligence.
| 0 | 0 | 0 | 1 | 0 | 0 |
Semi-Supervised Recurrent Neural Network for Adverse Drug Reaction Mention Extraction | Social media is an useful platform to share health-related information due to
its vast reach. This makes it a good candidate for public-health monitoring
tasks, specifically for pharmacovigilance. We study the problem of extraction
of Adverse-Drug-Reaction (ADR) mentions from social media, particularly from
twitter. Medical information extraction from social media is challenging,
mainly due to short and highly information nature of text, as compared to more
technical and formal medical reports.
Current methods in ADR mention extraction relies on supervised learning
methods, which suffers from labeled data scarcity problem. The State-of-the-art
method uses deep neural networks, specifically a class of Recurrent Neural
Network (RNN) which are Long-Short-Term-Memory networks (LSTMs)
\cite{hochreiter1997long}. Deep neural networks, due to their large number of
free parameters relies heavily on large annotated corpora for learning the end
task. But in real-world, it is hard to get large labeled data, mainly due to
heavy cost associated with manual annotation. Towards this end, we propose a
novel semi-supervised learning based RNN model, which can leverage unlabeled
data also present in abundance on social media. Through experiments we
demonstrate the effectiveness of our method, achieving state-of-the-art
performance in ADR mention extraction.
| 1 | 0 | 0 | 0 | 0 | 0 |
A Deep Neural Architecture for Sentence-level Sentiment Classification in Twitter Social Networking | This paper introduces a novel deep learning framework including a
lexicon-based approach for sentence-level prediction of sentiment label
distribution. We propose to first apply semantic rules and then use a Deep
Convolutional Neural Network (DeepCNN) for character-level embeddings in order
to increase information for word-level embedding. After that, a Bidirectional
Long Short-Term Memory Network (Bi-LSTM) produces a sentence-wide feature
representation from the word-level embedding. We evaluate our approach on three
Twitter sentiment classification datasets. Experimental results show that our
model can improve the classification accuracy of sentence-level sentiment
analysis in Twitter social networking.
| 1 | 0 | 0 | 0 | 0 | 0 |
A Tree-based Approach for Detecting Redundant Business Rules in very Large Financial Datasets | Net Asset Value (NAV) calculation and validation is the principle task of a
fund administrator. If the NAV of a fund is calculated incorrectly then there
is huge impact on the fund administrator; such as monetary compensation,
reputational loss, or loss of business. In general, these companies use the
same methodology to calculate the NAV of a fund, however the type of fund in
question dictates the set of business rules used to validate this. Today, most
Fund Administrators depend heavily on human resources due to the lack of an
automated standardized solutions, however due to economic climate and the need
for efficiency and costs reduction many banks are now looking for an automated
solution with minimal human interaction; i.e., straight through processing
(STP). Within the scope of a collaboration project that focuses on building an
optimal solution for NAV validation, in this paper, we will present a new
approach for detecting correlated business rules. We also show how we evaluate
this approach using real-world financial data.
| 1 | 0 | 0 | 0 | 0 | 0 |
Dynamics of the nonlinear Klein-Gordon equation in the nonrelativistic limit, I | The nonlinear Klein-Gordon (NLKG) equation on a manifold $M$ in the
nonrelativistic limit, namely as the speed of light $c$ tends to infinity, is
considered. In particular, a higher-order normalized approximation of NLKG
(which corresponds to the NLS at order $r=1$) is constructed, and when $M$ is a
smooth compact manifold or $\mathbb{R}^d$ it is proved that the solution of the
approximating equation approximates the solution of the NLKG locally uniformly
in time. When $M=\mathbb{R}^d$, $d \geq 3$, it is proved that solutions of the
linearized order $r$ normalized equation approximate solutions of linear
Klein-Gordon equation up to times of order $\mathcal{O}(c^{2(r-1)})$ for any
$r>1$.
| 0 | 0 | 1 | 0 | 0 | 0 |
Surface plasmons in superintense laser-solid interactions | We review studies of superintense laser interaction with solid targets where
the generation of propagating surface plasmons (or surface waves) plays a key
role. These studies include the onset of plasma instabilities at the irradiated
surface, the enhancement of secondary emissions (protons, electrons, and
photons as high harmonics in the XUV range) in femtosecond interactions with
grating targets, and the generation of unipolar current pulses with picosecond
duration. The experimental results give evidence of the existence of surface
plasmons in the nonlinear regime of relativistic electron dynamics. These
findings open up a route to the improvement of ultrashort laser-driven sources
of energetic radiation and, more in general, to the extension of plasmonics in
a high field regime.
| 0 | 1 | 0 | 0 | 0 | 0 |
Jacquard: A Large Scale Dataset for Robotic Grasp Detection | Grasping skill is a major ability that a wide number of real-life
applications require for robotisation. State-of-the-art robotic grasping
methods perform prediction of object grasp locations based on deep neural
networks. However, such networks require huge amount of labeled data for
training making this approach often impracticable in robotics. In this paper,
we propose a method to generate a large scale synthetic dataset with ground
truth, which we refer to as the Jacquard grasping dataset. Jacquard is built on
a subset of ShapeNet, a large CAD models dataset, and contains both RGB-D
images and annotations of successful grasping positions based on grasp attempts
performed in a simulated environment. We carried out experiments using an
off-the-shelf CNN, with three different evaluation metrics, including real
grasping robot trials. The results show that Jacquard enables much better
generalization skills than a human labeled dataset thanks to its diversity of
objects and grasping positions. For the purpose of reproducible research in
robotics, we are releasing along with the Jacquard dataset a web interface for
researchers to evaluate the successfulness of their grasping position
detections using our dataset.
| 1 | 0 | 0 | 0 | 0 | 0 |
Hochschild cohomology of some quantum complete intersections | We compute the Hochschild cohomology ring of the algebras $A= k\langle X,
Y\rangle/ (X^a, XY-qYX, Y^a)$ over a field $k$ where $a\geq 2$ and where $q\in
k$ is a primitive $a$-th root of unity. We find the the dimension of
$\mathrm{HH}^n(A)$ and show that it is independent of $a$. We compute
explicitly the ring structure of the even part of the Hochschild cohomology
modulo homogeneous nilpotent elements.
| 0 | 0 | 1 | 0 | 0 | 0 |
Stock Market Visualization | We provide complete source code for a front-end GUI and its back-end
counterpart for a stock market visualization tool. It is built based on the
"functional visualization" concept we discuss, whereby functionality is not
sacrificed for fancy graphics. The GUI, among other things, displays a
color-coded signal (computed by the back-end code) based on how "out-of-whack"
each stock is trading compared with its peers ("mean-reversion"), and the most
sizable changes in the signal ("momentum"). The GUI also allows to efficiently
filter/tier stocks by various parameters (e.g., sector, exchange, signal,
liquidity, market cap) and functionally display them. The tool can be run as a
web-based or local application.
| 0 | 0 | 0 | 0 | 0 | 1 |
Sharp estimates for solutions of mean field equation with collapsing singularity | The pioneering work of Brezis-Merle [7], Li-Shafrir [27], Li [26] and
Bartolucci-Tarantello [4] showed that any sequence of blow up solutions for
(singular) mean field equations of Liouville type must exhibit a "mass
concentration" property. A typical situation of blow-up occurs when we let the
singular (vortex) points involved in the equation (see (1.1) below) collapse
together. However in this case Lin-Tarantello in [30] pointed out that the
phenomenon: "bubbling implies mass concentration" might not occur and new
scenarios open for investigation. In this paper, we present two explicit
examples which illustrate (with mathematical rigor) how a "non-concentration"
situation does happen and its new features. Among other facts, we show that in
certain situations, the collapsing rate of the singularities can be used as
blow up parameter to describe the bubbling properties of the solution-sequence.
In this way we are able to establish accurate estimates around the blow-up
points which we hope to use towards a degree counting formula for the shadow
system (1.34) below.
| 0 | 0 | 1 | 0 | 0 | 0 |
Probabilistic Forwarding of Coded Packets on Networks | We consider a scenario of broadcasting information over a network of nodes
connected by noiseless communication links. A source node in the network has
$k$ data packets to broadcast, and it suffices that a large fraction of the
network nodes receives the broadcast. The source encodes the $k$ data packets
into $n \ge k$ coded packets using a maximum distance separable (MDS) code, and
transmits them to its one-hop neighbours. Every other node in the network
follows a probabilistic forwarding protocol, in which it forwards a previously
unreceived packet to all its neighbours with a certain probability $p$. A
"near-broadcast" is when the expected fraction of nodes that receive at least
$k$ of the $n$ coded packets is close to $1$. The forwarding probability $p$ is
chosen so as to minimize the expected total number of transmissions needed for
a near-broadcast. In this paper, we analyze the probabilistic forwarding of
coded packets on two specific network topologies: binary trees and square
grids. For trees, our analysis shows that for fixed $k$, the expected total
number of transmissions increases with $n$. On the other hand, on grids, we use
ideas from percolation theory to show that a judicious choice of $n$ will
significantly reduce the expected total number of transmissions needed for a
near-broadcast.
| 1 | 0 | 0 | 0 | 0 | 0 |
Quantifiers on languages and codensity monads | This paper contributes to the techniques of topo-algebraic recognition for
languages beyond the regular setting as they relate to logic on words. In
particular, we provide a general construction on recognisers corresponding to
adding one layer of various kinds of quantifiers and prove a related
Reutenauer-type theorem. Our main tools are codensity monads and duality
theory. Our construction hinges, in particular, on a measure-theoretic
characterisation of the profinite monad of the free S-semimodule monad for
finite and commutative semirings S, which generalises our earlier insight that
the Vietoris monad on Boolean spaces is the codensity monad of the finite
powerset functor.
| 1 | 0 | 1 | 0 | 0 | 0 |
Conservative Exploration using Interleaving | In many practical problems, a learning agent may want to learn the best
action in hindsight without ever taking a bad action, which is significantly
worse than the default production action. In general, this is impossible
because the agent has to explore unknown actions, some of which can be bad, to
learn better actions. However, when the actions are combinatorial, this may be
possible if the unknown action can be evaluated by interleaving it with the
production action. We formalize this concept as learning in stochastic
combinatorial semi-bandits with exchangeable actions. We design efficient
learning algorithms for this problem, bound their n-step regret, and evaluate
them on both synthetic and real-world problems. Our real-world experiments show
that our algorithms can learn to recommend K most attractive movies without
ever violating a strict production constraint, both overall and subject to a
diversity constraint.
| 0 | 0 | 0 | 1 | 0 | 0 |
Power Allocation for Full-Duplex Relay Selection in Underlay Cognitive Radio Networks: Coherent versus Non-Coherent Scenarios | This paper investigates power control and relay selection in Full Duplex
Cognitive Relay Networks (FDCRNs), where the secondary-user (SU) relays can
simultaneously receive data from the SU source and forward them to the SU
destination. We study both non-coherent and coherent scenarios. In the
non-coherent case, the SU relay forwards the signal from the SU source without
regulating the phase; while in the coherent scenario, the SU relay regulates
the phase when forwarding the signal to minimize the interference at the
primary-user (PU) receiver. We consider the problem of maximizing the
transmission rate from the SU source to the SU destination subject to the
interference constraint at the PU receiver and power constraints at both the SU
source and SU relay. We then develop a mathematical model to analyze the data
rate performance of the FDCRN considering the self-interference effects at the
FD relay. We develop low-complexity and high-performance joint power control
and relay selection algorithms. Extensive numerical results are presented to
illustrate the impacts of power level parameters and the self-interference
cancellation quality on the rate performance. Moreover, we demonstrate the
significant gain of phase regulation at the SU relay.
| 1 | 0 | 1 | 1 | 0 | 0 |
Carina: Interactive Million-Node Graph Visualization using Web Browser Technologies | We are working on a scalable, interactive visualization system, called
Carina, for people to explore million-node graphs. By using latest web browser
technologies, Carina offers fast graph rendering via WebGL, and works across
desktop (via Electron) and mobile platforms. Different from most existing graph
visualization tools, Carina does not store the full graph in RAM, enabling it
to work with graphs with up to 69M edges. We are working to improve and
open-source Carina, to offer researchers and practitioners a new, scalable way
to explore and visualize large graph datasets.
| 1 | 0 | 0 | 0 | 0 | 0 |
Domination between different products and finiteness of associated semi-norms | In this note we determine all possible dominations between different products
of manifolds, when none of the factors of the codomain is dominated by
products. As a consequence, we determine the finiteness of every
product-associated functorial semi-norm on the fundamental classes of the
aforementioned products. These results give partial answers to questions of M.
Gromov.
| 0 | 0 | 1 | 0 | 0 | 0 |
Discriminant of the ordinary transversal singularity type. The local aspects | Consider a space X with the singular locus, Z=Sing(X), of positive dimension.
Suppose both Z and X are locally complete intersections. The transversal type
of X along Z is generically constant but at some points of Z it degenerates. We
introduce (under certain conditions) the discriminant of the transversal type,
a subscheme of Z, that reflects these degenerations whenever the generic
transversal type is `ordinary'.
The scheme structure of this discriminant is imposed by various compatibility
properties and is often non-reduced. We establish the basic properties of this
discriminant: it is a Cartier divisor in Z, functorial under base change, flat
under some deformations of (X,Z), and compatible with pullback under some
morphisms, etc.
Furthermore, we study the local geometry of this discriminant, e.g. we
compute its multiplicity at a point, and we obtain the resolution of its
structure sheaf (as module on Z) and study the locally defining equation.
| 0 | 0 | 1 | 0 | 0 | 0 |
Possible spin excitation structure in monolayer FeSe grown on SrTiO$_{3}$ | Based on recent high-resolution angle-resolved photoemission spectroscopy
measurement in monolayer FeSe grown on SrTiO$_{3}$, we constructed a
tight-binding model and proposed a superconducting (SC) pairing function which
can well fit the observed band structure and SC gap anisotropy. Then we
investigated the spin excitation spectra in order to determine the possible
sign structure of the SC order parameter. We found that a resonance-like spin
excitation may occur if the SC order parameter changes sign along the Fermi
surfaces. However, this resonance is located at different locations in momentum
space compared to other FeSe-based superconductors, suggesting that the Fermi
surface shape and pairing symmetry in monolayer FeSe grown on SrTiO$_{3}$ may
be different from other FeSe-based superconductors.
| 0 | 1 | 0 | 0 | 0 | 0 |
Transient behavior of the solutions to the second order difference equations by the renormalization method based on Newton-Maclaurin expansion | The renormalization method based on the Newton-Maclaurin expansion is applied
to study the transient behavior of the solutions to the difference equations as
they tend to the steady-states. The key and also natural step is to make the
renormalization equations to be continuous such that the elementary functions
can be used to describe the transient behavior of the solutions to difference
equations. As the concrete examples, we deal with the important second order
nonlinear difference equations with a small parameter. The result shows that
the method is more natural than the multi-scale method.
| 0 | 1 | 1 | 0 | 0 | 0 |
Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting | We present a large-scale study of gender bias in occupation classification, a
task where the use of machine learning may lead to negative outcomes on
peoples' lives. We analyze the potential allocation harms that can result from
semantic representation bias. To do so, we study the impact on occupation
classification of including explicit gender indicators---such as first names
and pronouns---in different semantic representations of online biographies.
Additionally, we quantify the bias that remains when these indicators are
"scrubbed," and describe proxy behavior that occurs in the absence of explicit
gender indicators. As we demonstrate, differences in true positive rates
between genders are correlated with existing gender imbalances in occupations,
which may compound these imbalances.
| 1 | 0 | 0 | 1 | 0 | 0 |
SphereFace: Deep Hypersphere Embedding for Face Recognition | This paper addresses deep face recognition (FR) problem under open-set
protocol, where ideal face features are expected to have smaller maximal
intra-class distance than minimal inter-class distance under a suitably chosen
metric space. However, few existing algorithms can effectively achieve this
criterion. To this end, we propose the angular softmax (A-Softmax) loss that
enables convolutional neural networks (CNNs) to learn angularly discriminative
features. Geometrically, A-Softmax loss can be viewed as imposing
discriminative constraints on a hypersphere manifold, which intrinsically
matches the prior that faces also lie on a manifold. Moreover, the size of
angular margin can be quantitatively adjusted by a parameter $m$. We further
derive specific $m$ to approximate the ideal feature criterion. Extensive
analysis and experiments on Labeled Face in the Wild (LFW), Youtube Faces (YTF)
and MegaFace Challenge show the superiority of A-Softmax loss in FR tasks. The
code has also been made publicly available.
| 1 | 0 | 0 | 0 | 0 | 0 |
Speed-of-light pulses in the massless nonlinear Dirac equation with a potential | We consider the massless nonlinear Dirac (NLD) equation in $1+1$ dimension
with scalar-scalar self-interaction $\frac{g^2}{2} (\bar{\Psi} \Psi)^2$ in the
presence of three external electromagnetic potentials $V(x)$, a potential
barrier, a constant potential, and a potential well. By solving numerically the
NLD equation, we find that, for all three cases, after a short transit time,
the initial pulse breaks into two pulses which are solutions of the massless
linear Dirac equation traveling in opposite directions with the speed of light.
During this splitting the charge and the energy are conserved, whereas the
momentum is conserved when the solutions possess specific symmetries. For the
case of the constant potential, we derive exact analytical solutions of the
massless NLD equation that are also solutions of the massless linearized Dirac
equation.
| 0 | 1 | 0 | 0 | 0 | 0 |
A Hybrid Feasibility Constraints-Guided Search to the Two-Dimensional Bin Packing Problem with Due Dates | The two-dimensional non-oriented bin packing problem with due dates packs a
set of rectangular items, which may be rotated by 90 degrees, into identical
rectangular bins. The bins have equal processing times. An item's lateness is
the difference between its due date and the completion time of its bin. The
problem packs all items without overlap as to minimize maximum lateness Lmax.
The paper proposes a tight lower bound that enhances an existing bound on
Lmax for 24.07% of the benchmark instances and matches it in 30.87% cases. In
addition, it models the problem using mixed integer programming (MIP), and
solves small-sized instances exactly using CPLEX. It approximately solves
larger-sized instances using a two-stage heuristic. The first stage constructs
an initial solution via a first-fit heuristic that applies an iterative
constraint programming (CP)-based neighborhood search. The second stage, which
is iterative too, approximately solves a series of assignment low-level MIPs
that are guided by feasibility constraints. It then enhances the solution via a
high-level random local search. The approximate approach improves existing
upper bounds by 27.45% on average, and obtains the optimum for 33.93% of the
instances. Overall, the exact and approximate approaches identify the optimum
for 39.07% cases.
The proposed approach is applicable to complex problems. It applies CP and
MIP sequentially, while exploring their advantages, and hybridizes heuristic
search with MIP. It embeds a new lookahead strategy that guards against
infeasible search directions and constrains the search to improving directions
only; thus, differs from traditional lookahead beam searches.
| 1 | 0 | 0 | 0 | 0 | 0 |
The earliest phases of high-mass star formation, as seen in NGC 6334 by \emph{Herschel} | To constrain models of high-mass star formation, the Herschel/HOBYS KP aims
at discovering massive dense cores (MDCs) able to host the high-mass analogs of
low-mass prestellar cores, which have been searched for over the past decade.
We here focus on NGC6334, one of the best-studied HOBYS molecular cloud
complexes.
We used Herschel PACS and SPIRE 70-500mu images of the NGC6334 complex
complemented with (sub)millimeter and mid-infrared data. We built a complete
procedure to extract ~0.1 pc dense cores with the getsources software, which
simultaneously measures their far-infrared to millimeter fluxes. We carefully
estimated the temperatures and masses of these dense cores from their SEDs.
A cross-correlation with high-mass star formation signposts suggests a mass
threshold of 75Msun for MDCs in NGC6334. MDCs have temperatures of 9.5-40K,
masses of 75-1000Msun, and densities of 10^5-10^8cm-3. Their mid-IR emission is
used to separate 6 IR-bright and 10 IR-quiet protostellar MDCs while their 70mu
emission strength, with respect to fitted SEDs, helps identify 16 starless MDC
candidates. The ability of the latter to host high-mass prestellar cores is
investigated here and remains questionable. An increase in mass and density
from the starless to the IR-quiet and IR-bright phases suggests that the
protostars and MDCs simultaneously grow in mass. The statistical lifetimes of
the high-mass prestellar and protostellar core phases, estimated to be
1-7x10^4yr and at most 3x10^5yr respectively, suggest a dynamical scenario of
high-mass star formation.
The present study provides good mass estimates for a statistically
significant sample, covering the earliest phases of high-mass star formation.
High-mass prestellar cores may not exist in NGC6334, favoring a scenario
presented here, which simultaneously forms clouds and high-mass protostars.
| 0 | 1 | 0 | 0 | 0 | 0 |
Robust consistent a posteriori error majorants for approximate solutions of diffusion-reaction equations | Efficiency of the error control of numerical solutions of partial
differential equations entirely depends on the two factors: accuracy of an a
posteriori error majorant and the computational cost of its evaluation for some
test function/vector-function plus the cost of the latter. In the paper,
consistency of an a posteriori bound implies that it is the same in the order
with the respective unimprovable a priori bound. Therefore, it is the basic
characteristic related to the first factor. The paper is dedicated to the
elliptic diffusion-reaction equations. We present a guaranteed robust a
posteriori error majorant effective at any nonnegative constant reaction
coefficient (r.c.). For a wide range of finite element solutions on a
quasiuniform meshes the majorant is consistent. For big values of r.c. the
majorant coincides with the majorant of Aubin (1972), which, as it is known,
for not big r.c. ($<ch^{-2}$) is inconsistent and loses its sense at r.c.
approaching zero. Our majorant improves also some other majorants derived for
the Poisson and reaction-diffusion equations.
| 0 | 0 | 1 | 0 | 0 | 0 |
Extended opportunity cost model to find near equilibrium electricity prices under non-convexities | This paper finds near equilibrium prices for electricity markets with
nonconvexities due to binary variables, in order to reduce the market
participants' opportunity costs, such as generators' unrecovered costs. The
opportunity cost is defined as the difference between the profit when the
instructions of the market operator are followed and when the market
participants can freely make their own decisions based on the market prices. We
use the minimum complementarity approximation to the minimum total opportunity
cost (MTOC) model, from previous research, with tests on a much more realistic
unit commitment (UC) model than in previous research, including features such
as reserve requirements, ramping constraints, and minimum up and down times.
The developed model incorporates flexible price responsive demand, as in
previous research, but since not all demand is price responsive, we consider
the more realistic case that total demand is a mixture of fixed and flexible.
Another improvement over previous MTOC research is computational: whereas the
previous research had nonconvex terms among the objective function's continuous
variables, we convert the objective to an equivalent form that contains only
linear and convex quadratic terms in the continuous variables. We compare the
unit commitment model with the standard social welfare optimization version of
UC, in a series of sensitivity analyses, varying flexible demand to represent
varying degrees of future penetration of electric vehicles and smart
appliances, different ratios of generation availability, and different values
of transmission line capacities to consider possible congestion. The minimum
total opportunity cost and social welfare solutions are mostly very close in
different scenarios, except in some extreme cases.
| 0 | 0 | 0 | 0 | 0 | 1 |
Kernel Two-Sample Hypothesis Testing Using Kernel Set Classification | The two-sample hypothesis testing problem is studied for the challenging
scenario of high dimensional data sets with small sample sizes. We show that
the two-sample hypothesis testing problem can be posed as a one-class set
classification problem. In the set classification problem the goal is to
classify a set of data points that are assumed to have a common class. We prove
that the average probability of error given a set is less than or equal to the
Bayes error and decreases as a power of $n$ number of sample data points in the
set. We use the positive definite Set Kernel for directly mapping sets of data
to an associated Reproducing Kernel Hilbert Space, without the need to learn a
probability distribution. We specifically solve the two-sample hypothesis
testing problem using a one-class SVM in conjunction with the proposed Set
Kernel. We compare the proposed method with the Maximum Mean Discrepancy,
F-Test and T-Test methods on a number of challenging simulated high dimensional
and small sample size data. We also perform two-sample hypothesis testing
experiments on six cancer gene expression data sets and achieve zero type-I and
type-II error results on all data sets.
| 0 | 0 | 0 | 1 | 0 | 0 |
An Efficient Approach for Removing Look-ahead Bias in the Least Square Monte Carlo Algorithm: Leave-One-Out | The least square Monte Carlo (LSM) algorithm proposed by Longstaff and
Schwartz [2001] is the most widely used method for pricing options with early
exercise features. The LSM estimator contains look-ahead bias, and the
conventional technique of removing it necessitates an independent set of
simulations. This study proposes a new approach for efficiently eliminating
look-ahead bias by using the leave-one-out method, a well-known
cross-validation technique for machine learning applications. The leave-one-out
LSM (LOOLSM) method is illustrated with examples, including multi-asset options
whose LSM price is biased high. The asymptotic behavior of look-ahead bias is
also discussed with the LOOLSM approach.
| 0 | 0 | 0 | 0 | 0 | 1 |
Accurately and Efficiently Interpreting Human-Robot Instructions of Varying Granularities | Humans can ground natural language commands to tasks at both abstract and
fine-grained levels of specificity. For instance, a human forklift operator can
be instructed to perform a high-level action, like "grab a pallet" or a
low-level action like "tilt back a little bit." While robots are also capable
of grounding language commands to tasks, previous methods implicitly assume
that all commands and tasks reside at a single, fixed level of abstraction.
Additionally, methods that do not use multiple levels of abstraction encounter
inefficient planning and execution times as they solve tasks at a single level
of abstraction with large, intractable state-action spaces closely resembling
real world complexity. In this work, by grounding commands to all the tasks or
subtasks available in a hierarchical planning framework, we arrive at a model
capable of interpreting language at multiple levels of specificity ranging from
coarse to more granular. We show that the accuracy of the grounding procedure
is improved when simultaneously inferring the degree of abstraction in language
used to communicate the task. Leveraging hierarchy also improves efficiency:
our proposed approach enables a robot to respond to a command within one second
on 90% of our tasks, while baselines take over twenty seconds on half the
tasks. Finally, we demonstrate that a real, physical robot can ground commands
at multiple levels of abstraction allowing it to efficiently plan different
subtasks within the same planning hierarchy.
| 1 | 0 | 0 | 0 | 0 | 0 |
Nonlocal Venttsel' diffusion in fractal-type domains: regularity results and numerical approximation | We study a nonlocal Venttsel' problem in a non-convex bounded domain with a
Koch-type boundary. Regularity results of the strict solution are proved in
weighted Sobolev spaces. The numerical approximation of the problem is carried
out and optimal a priori error estimates are obtained.
| 0 | 0 | 1 | 0 | 0 | 0 |
Magnon Spin-Momentum Locking: Various Spin Vortices and Dirac Magnons in Noncollinear Antiferromagnets | We generalize the concept of the spin-momentum locking to magnonic systems
and derive the formula to calculate the spin expectation value for one-magnon
states of general two-body spin Hamiltonians. We give no-go conditions for
magnon spin to be independent of momentum. As examples of the magnon
spin-momentum locking, we analyze a one-dimensional antiferromagnet with the
Néel order and two-dimensional kagome lattice antiferromagnets with the
120$^\circ$ structure. We find that the magnon spin depends on its momentum
even when the Hamiltonian has the $z$-axis spin rotational symmetry, which can
be explained in the context of a singular band point or a $U(1)$ symmetry
breaking. A spin vortex in momentum space generated in a kagome lattice
antiferromagnet has the winding number $Q=-2$, while the typical one observed
in topological insulator surface states is characterized by $Q=+1$. A magnonic
analogue of the surface states, the Dirac magnon with $Q=+1$, is found in
another kagome lattice antiferromagnet. We also derive the sum rule for $Q$ by
using the Poincaré-Hopf index theorem.
| 0 | 1 | 0 | 0 | 0 | 0 |
Analytic evaluation of some three- and four- electron atomic integrals involving s STO's and exponential correlation with unlinked $r_{ij}$'s | The method of evaluation outlined in a previous work has been utilized here
to evaluate certain other three- electron and four- electron atomic integrals
involving s Slater-type orbitals and exponential correlation with unlinked
$r_{ij}$'s. Limiting expressions for various such integrals have been derived,
which has not been done earlier. Closed-form expressions for $<r_{12} r_{13} /
r_{14}>$, $<r_{12}r_{34}/r_{23}>$, $<r_{12}r_{23}/r_{34}>$,
$<r_{12}r_{13}/r_{34}>$ and $<r_{12}r_{34}/r_{13}>$ have been obtained.
| 0 | 1 | 0 | 0 | 0 | 0 |
GIANT: Globally Improved Approximate Newton Method for Distributed Optimization | For distributed computing environment, we consider the empirical risk
minimization problem and propose a distributed and communication-efficient
Newton-type optimization method. At every iteration, each worker locally finds
an Approximate NewTon (ANT) direction, which is sent to the main driver. The
main driver, then, averages all the ANT directions received from workers to
form a {\it Globally Improved ANT} (GIANT) direction. GIANT is highly
communication efficient and naturally exploits the trade-offs between local
computations and global communications in that more local computations result
in fewer overall rounds of communications. Theoretically, we show that GIANT
enjoys an improved convergence rate as compared with first-order methods and
existing distributed Newton-type methods. Further, and in sharp contrast with
many existing distributed Newton-type methods, as well as popular first-order
methods, a highly advantageous practical feature of GIANT is that it only
involves one tuning parameter. We conduct large-scale experiments on a computer
cluster and, empirically, demonstrate the superior performance of GIANT.
| 1 | 0 | 0 | 1 | 0 | 0 |
Stability and performance analysis of linear positive systems with delays using input-output methods | It is known that input-output approaches based on scaled small-gain theorems
with constant $D$-scalings and integral linear constraints are non-conservative
for the analysis of some classes of linear positive systems interconnected with
uncertain linear operators. This dramatically contrasts with the case of
general linear systems with delays where input-output approaches provide, in
general, sufficient conditions only. Using these results we provide simple
alternative proofs for many of the existing results on the stability of linear
positive systems with discrete/distributed/neutral time-invariant/-varying
delays and linear difference equations. In particular, we give a simple proof
for the characterization of diagonal Riccati stability for systems with
discrete-delays and generalize this equation to other types of delay systems.
The fact that all those results can be reproved in a very simple way
demonstrates the importance and the efficiency of the input-output framework
for the analysis of linear positive systems. The approach is also used to
derive performance results evaluated in terms of the $L_1$-, $L_2$- and
$L_\infty$-gains. It is also flexible enough to be used for design purposes.
| 1 | 0 | 1 | 0 | 0 | 0 |
Joint Inference of User Community and Interest Patterns in Social Interaction Networks | Online social media have become an integral part of our social beings.
Analyzing conversations in social media platforms can lead to complex
probabilistic models to understand social interaction networks. In this paper,
we present a modeling approach for characterizing social interaction networks
by jointly inferring user communities and interests based on social media
interactions. We present several pattern inference models: i) Interest pattern
model (IPM) captures population level interaction topics, ii) User interest
pattern model (UIPM) captures user specific interaction topics, and iii)
Community interest pattern model (CIPM) captures both community structures and
user interests. We test our methods on Twitter data collected from Purdue
University community. From our model results, we observe the interaction topics
and communities related to two big events within Purdue University community,
namely Purdue Day of Giving and Senator Bernie Sanders' visit to Purdue
University as part of Indiana Primary Election 2016. Constructing social
interaction networks based on user interactions accounts for the similarity of
users' interactions on various topics of interest and indicates their community
belonging further beyond connectivity. We observed that the
degree-distributions of such networks follow power-law that is indicative of
the existence of fewer nodes in the network with higher levels of interactions,
and many other nodes with less interactions. We also discuss the application of
such networks as a useful tool to effectively disseminate specific information
to the target audience towards planning any large-scale events and demonstrate
how to single out specific nodes in a given community by running network
algorithms.
| 1 | 1 | 0 | 0 | 0 | 0 |
A mode theory for the electoweak interaction and its application to neutrino masses | A theory is proposed, in which the basic elements of reality are assumed to
be something called modes. Particles are interpreted as composites of modes,
corresponding to eigenstates of the interaction Hamiltonian of modes. At the
fundamental level of the proposed theory, there are two basic modes only,whose
spinor spaces are the two smallest nontrivial representation spaces of the
SL(2,C) group, one being the complex conjugate of the other. All other modes
are constructed from the two basic modes, making use of the operations of
direct sum and direct product for related spinor spaces. Accompanying the
construction of direct-product modes, interactions among modes are introduced
in a natural way, with the interaction Hamiltonian given from mappings between
the corresponding state spaces. The interaction Hamiltonian thus obtained turn
out to possess a form, which is similar to a major part of the interaction
Hamiltonian in the Glashow-Weinberg-Salam electroweak theory. In the proposed
theory, it is possible for the second-order perturbation expansion of energy to
be free from ultraviolet divergence. This expansion is used to derive some
approximate relations for neutrino masses; in particular, a rough estimate is
obtained for the ratio of mass differences of neutrinos, which gives the
correct order of magnitude compared with the experimental result.
| 0 | 1 | 0 | 0 | 0 | 0 |
Guided Machine Learning for power grid segmentation | The segmentation of large scale power grids into zones is crucial for control
room operators when managing the grid complexity near real time. In this paper
we propose a new method in two steps which is able to automatically do this
segmentation, while taking into account the real time context, in order to help
them handle shifting dynamics. Our method relies on a "guided" machine learning
approach. As a first step, we define and compute a task specific "Influence
Graph" in a guided manner. We indeed simulate on a grid state chosen
interventions, representative of our task of interest (managing active power
flows in our case). For visualization and interpretation, we then build a
higher representation of the grid relevant to this task by applying the graph
community detection algorithm \textit{Infomap} on this Influence Graph. To
illustrate our method and demonstrate its practical interest, we apply it on
commonly used systems, the IEEE-14 and IEEE-118. We show promising and original
interpretable results, especially on the previously well studied RTS-96 system
for grid segmentation. We eventually share initial investigation and results on
a large-scale system, the French power grid, whose segmentation had a
surprising resemblance with RTE's historical partitioning.
| 0 | 0 | 0 | 1 | 0 | 0 |
121,123Sb NQR as a microscopic probe in Te doped correlated semimetal FeSb2 : emergence of electronic Griffith phase, magnetism and metallic behavior % | $^{121,123}Sb$ nuclear quadrupole resonance (NQR) was applied to
$Fe(Sb_{1-x}Te_x)_2$ in the low doping regime (\emph{x = 0, 0.01} and
\emph{0.05}) as a microscopic zero field probe to study the evolution of
\emph{3d} magnetism and the emergence of metallic behavior. Whereas the NQR
spectra itself reflects the degree of local disorder via the width of the
individual NQR lines, the spin lattice relaxation rate (SLRR) $1/T_1(T)$ probes
the fluctuations at the $Sb$ - site. The fluctuations originate either from
conduction electrons or from magnetic moments. In contrast to the semi metal
$FeSb_2$ with a clear signature of the charge and spin gap formation in
$1/T_1(T)T ( \sim exp/ (\Delta k_BT) ) $, the 1\% $Te$ doped system exhibits
almost metallic conductivity and a almost filled gap. A weak divergence of the
SLRR coefficient $1/T_1(T)T \sim T^{-n} \sim T^{-0.2}$ points towards the
presence of electronic correlations towards low temperatures wheras the
\textit{5\%} $Te$ doped sample exhibits a much larger divergence in the SLRR
coefficient showing $1/T_1(T)T \sim T^{-0.72} $. According to the specific heat
divergence a power law with $n\ =\ 2\ m\ =\ 0.56$ is expected for the SLRR.
Furthermore $Te$-doped $FeSb_2$ as a disordered paramagnetic metal might be a
platform for the electronic Griffith phase scenario. NQR evidences a
substantial asymmetric broadening of the $^{121,123}Sb$ NQR spectrum for the
\emph{5\%} sample. This has purely electronic origin in agreement with the
electronic Griffith phase and stems probably from an enhanced $Sb$-$Te$ bond
polarization and electronic density shift towards the $Te$ atom inside
$Sb$-$Te$ dumbbell.
| 0 | 1 | 0 | 0 | 0 | 0 |
The Merging Path Plot: adaptive fusing of k-groups with likelihood-based model selection | There are many statistical tests that verify the null hypothesis: the
variable of interest has the same distribution among k-groups. But once the
null hypothesis is rejected, how to present the structure of dissimilarity
between groups? In this article, we introduce The Merging Path Plot - a
methodology, and factorMerger - an R package, for exploration and visualization
of k-group dissimilarities. Comparison of k-groups is one of the most important
issues in exploratory analyses and it has zillions of applications. The
classical solution is to test a~null hypothesis that observations from all
groups come from the same distribution. If the global null hypothesis is
rejected, a~more detailed analysis of differences among pairs of groups is
performed. The traditional approach is to use pairwise post hoc tests in order
to verify which groups differ significantly. However, this approach fails with
a large number of groups in both interpretation and visualization layer.
The~Merging Path Plot methodology solves this problem by using an
easy-to-understand description of dissimilarity among groups based on
Likelihood Ratio Test (LRT) statistic.
| 1 | 0 | 0 | 1 | 0 | 0 |
Constraint on cosmological parameters by Hubble parameter from gravitational wave standard sirens of neutron star binary system | In this paper, we present a new method of measuring Hubble parameter($H(z)$),
making use of the anisotropy of luminosity distance($d_{L}$), and the analysis
of gravitational wave(GW) of neutron star(NS) binary system. The method has
never been put into practice before due to the lack of the ability of detecting
GW. LIGO's success in detecting GW of black hole(BH) binary system merger
announced the possibility of this new method. We apply this method to several
GW detecting projects, including Advanced LIGO(Adv-LIGO), Einstein
Telescope(ET) and DECIGO, finding that the $H(z)$ by Adv-LIGO and ET is of bad
accuracy, while the $H(z)$ by DECIGO shows a good accuracy. We use the error
information of $H(z)$ by DECIGO to simulate $H(z)$ data at every 0.1 redshift
span, and put the mock data into the forecasting of cosmological parameters.
Compared with the available 38 observed $H(z)$ data(OHD), mock data shows an
obviously tighter constraint on cosmological parameters, and a concomitantly
higher value of Figure of Merit(FoM). For a 3-year-observation by standard
sirens of DECIGO, the FoM value is as high as 834.9. If a 10-year-observation
is launched, the FoM could reach 2783.1. For comparison, the FoM of 38 actual
observed $H(z)$ data is 9.3. These improvement indicates that the new method
has great potential in further cosmological constraints.
| 0 | 1 | 0 | 0 | 0 | 0 |
Semidefinite tests for latent causal structures | Testing whether a probability distribution is compatible with a given
Bayesian network is a fundamental task in the field of causal inference, where
Bayesian networks model causal relations. Here we consider the class of causal
structures where all correlations between observed quantities are solely due to
the influence from latent variables. We show that each model of this type
imposes a certain signature on the observable covariance matrix in terms of a
particular decomposition into positive semidefinite components. This signature,
and thus the underlying hypothetical latent structure, can be tested in a
computationally efficient manner via semidefinite programming. This stands in
stark contrast with the algebraic geometric tools required if the full
observable probability distribution is taken into account. The semidefinite
test is compared with tests based on entropic inequalities.
| 0 | 0 | 1 | 1 | 0 | 0 |
From atomistic model to the Peierls-Nabarro model with $γ$-surface for dislocations | The Peierls-Nabarro (PN) model for dislocations is a hybrid model that
incorporates the atomistic information of the dislocation core structure into
the continuum theory. In this paper, we study the convergence from a full
atomistic model to the PN model with $\gamma$-surface for the dislocation in a
bilayer system (e.g. bilayer graphene). We prove that the displacement field of
and the total energy of the dislocation solution of the PN model are
asymptotically close to those of the full atomistic model. Our work can be
considered as a generalization of the analysis of the convergence from
atomistic model to Cauchy-Born rule for crystals without defects in the
literature.
| 0 | 0 | 1 | 0 | 0 | 0 |
A family of monogenic $S_4$ quartic fields arising from elliptic curves | We consider partial torsion fields (fields generated by a root of a division
polynomial) for elliptic curves. By analysing the reduction properties of
elliptic curves, and applying the Montes Algorithm, we obtain information about
the ring of integers. In particular, for the partial $3$-torsion fields for a
certain one-parameter family of non-CM elliptic curves, we describe a power
basis. As a result, we show that the one-parameter family of quartic $S_4$
fields given by $T^4 - 6T^2 - \alpha T - 3$ for $\alpha \in \mathbb{Z}$ such
that $\alpha \pm 8$ are squarefree, are monogenic.
| 0 | 0 | 1 | 0 | 0 | 0 |
Aggregation and Resource Scheduling in Machine-type Communication Networks: A Stochastic Geometry Approach | Data aggregation is a promising approach to enable massive machine-type
communication (mMTC). This paper focuses on the aggregation phase where a
massive number of machine-type devices (MTDs) transmit to aggregators. By using
non-orthogonal multiple access (NOMA) principles, we allow several MTDs to
share the same orthogonal channel in our proposed hybrid access scheme. We
develop an analytical framework based on stochastic geometry to investigate the
system performance in terms of average success probability and average number
of simultaneously served MTDs, under imperfect successive interference
cancellation (SIC) at the aggregators, for two scheduling schemes: random
resource scheduling (RRS) and channel-aware resource scheduling (CRS). We
identify the power constraints on the MTDs sharing the same channel to attain a
fair coexistence with purely orthogonal multiple access (OMA) setups, then
power control coefficients are found so that these MTDs perform with similar
reliability. We show that under high access demand, the hybrid scheme with CRS
outperforms the OMA setup by simultaneously serving more MTDs with reduced
power consumption.
| 1 | 0 | 0 | 1 | 0 | 0 |
Setting Players' Behaviors in World of Warcraft through Semi-Supervised Learning | Digital games are one of the major and most important fields on the
entertainment domain, which also involves cinema and music. Numerous attempts
have been done to improve the quality of the games including more realistic
artistic production and computer science. Assessing the player's behavior, a
task known as player modeling, is currently the need of the hour which leads to
possible improvements in terms of: (i) better game interaction experience, (ii)
better exploitation of the relationship between players, and (iii)
increasing/maintaining the number of players interested in the game. In this
paper we model players using the basic four behaviors proposed in
\cite{BartleArtigo}, namely: achiever, explorer, socializer and killer. Our
analysis is carried out using data obtained from the game "World of Warcraft"
over 3 years (2006 $-$ 2009). We employ a semi-supervised learning technique in
order to find out characteristics that possibly impact player's behavior.
| 1 | 0 | 0 | 0 | 0 | 0 |
Boundary Hamiltonian theory for gapped topological orders | In this letter, we report our systematic construction of the lattice
Hamiltonian model of topological orders on open surfaces, with explicit
boundary terms. We do this mainly for the Levin-Wen stringnet model. The full
Hamiltonian in our approach yields a topologically protected, gapped energy
spectrum, with the corresponding wave functions robust under
topology-preserving transformations of the lattice of the system. We explicitly
present the wavefunctions of the ground states and boundary elementary
excitations. We construct the creation and hopping operators of boundary
quasi-particles. We find that given a bulk topological order, the gapped
boundary conditions are classified by Frobenius algebras in its input data.
Emergent topological properties of the ground states and boundary excitations
are characterized by (bi-) modules over Frobenius algebras.
| 0 | 1 | 1 | 0 | 0 | 0 |
Euler characteristics of cominuscule quantum K-theory | We prove an identity relating the product of two opposite Schubert varieties
in the (equivariant) quantum K-theory ring of a cominuscule flag variety to the
minimal degree of a rational curve connecting the Schubert varieties. We deduce
that the sum of the structure constants associated to any product of Schubert
classes is equal to one. Equivalently, the sheaf Euler characteristic map
extends to a ring homomorphism defined on the quantum K-theory ring.
| 0 | 0 | 1 | 0 | 0 | 0 |
Dialogue Act Sequence Labeling using Hierarchical encoder with CRF | Dialogue Act recognition associate dialogue acts (i.e., semantic labels) to
utterances in a conversation. The problem of associating semantic labels to
utterances can be treated as a sequence labeling problem. In this work, we
build a hierarchical recurrent neural network using bidirectional LSTM as a
base unit and the conditional random field (CRF) as the top layer to classify
each utterance into its corresponding dialogue act. The hierarchical network
learns representations at multiple levels, i.e., word level, utterance level,
and conversation level. The conversation level representations are input to the
CRF layer, which takes into account not only all previous utterances but also
their dialogue acts, thus modeling the dependency among both, labels and
utterances, an important consideration of natural dialogue. We validate our
approach on two different benchmark data sets, Switchboard and Meeting Recorder
Dialogue Act, and show performance improvement over the state-of-the-art
methods by $2.2\%$ and $4.1\%$ absolute points, respectively. It is worth
noting that the inter-annotator agreement on Switchboard data set is $84\%$,
and our method is able to achieve the accuracy of about $79\%$ despite being
trained on the noisy data.
| 1 | 0 | 0 | 0 | 0 | 0 |
Noise Stability is computable and low dimensional | Questions of noise stability play an important role in hardness of
approximation in computer science as well as in the theory of voting. In many
applications, the goal is to find an optimizer of noise stability among all
possible partitions of $\mathbb{R}^n$ for $n \geq 1$ to $k$ parts with given
Gaussian measures $\mu_1,\ldots,\mu_k$. We call a partition $\epsilon$-optimal,
if its noise stability is optimal up to an additive $\epsilon$. In this paper,
we give an explicit, computable function $n(\epsilon)$ such that an
$\epsilon$-optimal partition exists in $\mathbb{R}^{n(\epsilon)}$. This result
has implications for the computability of certain problems in non-interactive
simulation, which are addressed in a subsequent work.
| 1 | 0 | 1 | 0 | 0 | 0 |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.