ID
int64 1
21k
| TITLE
stringlengths 7
239
| ABSTRACT
stringlengths 7
2.76k
| Computer Science
int64 0
1
| Physics
int64 0
1
| Mathematics
int64 0
1
| Statistics
int64 0
1
| Quantitative Biology
int64 0
1
| Quantitative Finance
int64 0
1
|
|---|---|---|---|---|---|---|---|---|
20,201 | Pitfalls and Best Practices in Algorithm Configuration | Good parameter settings are crucial to achieve high performance in many areas
of artificial intelligence (AI), such as propositional satisfiability solving,
AI planning, scheduling, and machine learning (in particular deep learning).
Automated algorithm configuration methods have recently received much attention
in the AI community since they replace tedious, irreproducible and error-prone
manual parameter tuning and can lead to new state-of-the-art performance.
However, practical applications of algorithm configuration are prone to several
(often subtle) pitfalls in the experimental design that can render the
procedure ineffective. We identify several common issues and propose best
practices for avoiding them. As one possibility for automatically handling as
many of these as possible, we also propose a tool called GenericWrapper4AC.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,202 | Critical behaviors in contagion dynamics | We study the critical behavior of a general contagion model where nodes are
either active (e.g. with opinion A, or functioning) or inactive (e.g. with
opinion B, or damaged). The transitions between these two states are determined
by (i) spontaneous transitions independent of the neighborhood, (ii)
transitions induced by neighboring nodes and (iii) spontaneous reverse
transitions. The resulting dynamics is extremely rich including limit cycles
and random phase switching. We derive a unifying mean-field theory.
Specifically, we analytically show that the critical behavior of systems whose
dynamics is governed by processes (i-iii) can only exhibit three distinct
regimes: (a) uncorrelated spontaneous transition dynamics (b) contact process
dynamics and (c) cusp catastrophes. This ends a long-standing debate on the
universality classes of complex contagion dynamics in mean-field and
substantially deepens its mathematical understanding.
| 0 | 1 | 1 | 0 | 0 | 0 |
20,203 | Machine Learning in Appearance-based Robot Self-localization | An appearance-based robot self-localization problem is considered in the
machine learning framework. The appearance space is composed of all possible
images, which can be captured by a robot's visual system under all robot
localizations. Using recent manifold learning and deep learning techniques, we
propose a new geometrically motivated solution based on training data
consisting of a finite set of images captured in known locations of the robot.
The solution includes estimation of the robot localization mapping from the
appearance space to the robot localization space, as well as estimation of the
inverse mapping for modeling visual image features. The latter allows solving
the robot localization problem as the Kalman filtering problem.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,204 | Towards Communication-Aware Robust Topologies | We currently witness the emergence of interesting new network topologies
optimized towards the traffic matrices they serve, such as demand-aware
datacenter interconnects (e.g., ProjecToR) and demand-aware overlay networks
(e.g., SplayNets). This paper introduces a formal framework and approach to
reason about and design such topologies. We leverage a connection between the
communication frequency of two nodes and the path length between them in the
network, which depends on the entropy of the communication matrix. Our main
contribution is a novel robust, yet sparse, family of network topologies which
guarantee an expected path length that is proportional to the entropy of the
communication patterns.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,205 | Zero-shot Domain Adaptation without Domain Semantic Descriptors | We propose a method to infer domain-specific models such as classifiers for
unseen domains, from which no data are given in the training phase, without
domain semantic descriptors. When training and test distributions are
different, standard supervised learning methods perform poorly. Zero-shot
domain adaptation attempts to alleviate this problem by inferring models that
generalize well to unseen domains by using training data in multiple source
domains. Existing methods use observed semantic descriptors characterizing
domains such as time information to infer the domain-specific models for the
unseen domains. However, it cannot always be assumed that such metadata can be
used in real-world applications. The proposed method can infer appropriate
domain-specific models without any semantic descriptors by introducing the
concept of latent domain vectors, which are latent representations for the
domains and are used for inferring the models. The latent domain vector for the
unseen domain is inferred from the set of the feature vectors in the
corresponding domain, which is given in the testing phase. The domain-specific
models consist of two components: the first is for extracting a representation
of a feature vector to be predicted, and the second is for inferring model
parameters given the latent domain vector. The posterior distributions of the
latent domain vectors and the domain-specific models are parametrized by neural
networks, and are optimized by maximizing the variational lower bound using
stochastic gradient descent. The effectiveness of the proposed method was
demonstrated through experiments using one regression and two classification
tasks.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,206 | Network modelling of topological domains using Hi-C data | Genome-wide chromosome conformation capture techniques such as Hi-C enable
the generation of 3D genome contact maps and offer new pathways toward
understanding the spatial organization of genome. One specific feature of the
3D organization is known as topologically associating domains (TADs), which are
densely interacting, contiguous chromatin regions playing important roles in
regulating gene expression. A few algorithms have been proposed to detect TADs.
In particular, the structure of Hi-C data naturally inspires application of
community detection methods. However, one of the drawbacks of community
detection is that most methods take exchangeability of the nodes in the network
for granted; whereas the nodes in this case, i.e. the positions on the
chromosomes, are not exchangeable. We propose a network model for detecting
TADs using Hi-C data that takes into account this non-exchangeability. In
addition, our model explicitly makes use of cell-type specific CTCF binding
sites as biological covariates and can be used to identify conserved TADs
across multiple cell types. The model leads to a likelihood objective that can
be efficiently optimized via relaxation. We also prove that when suitably
initialized, this model finds the underlying TAD structure with high
probability. Using simulated data, we show the advantages of our method and the
caveats of popular community detection methods, such as spectral clustering, in
this application. Applying our method to real Hi-C data, we demonstrate the
domains identified have desirable epigenetic features and compare them across
different cell types. The code is available upon request.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,207 | On a Possible Giant Impact Origin for the Colorado Plateau | It is proposed and substantiated that an extraterrestrial object of the
approximate size and mass of Planet Mars, impacting the Earth in an oblique
angle along an approximately NE-SW route (with respect to the current
orientation of the North America continent) around 750 million years ago (750
Ma), is likely to be the direct cause of a chain of events which led to the
rifting of the Rodinia supercontinent and the severing of the foundation of the
Colorado Plateau from its surrounding craton.
It is further argued that the impactor most likely originated as a rouge
exoplanet produced during one of the past crossings of our Solar System through
the Galactic spiral arms in its orbital motion around the center of the Milky
Way Galaxy. Recent work has shown that the sites of galactic spiral arms are
locations of density-wave collisionless shocks. The perturbations from such
shock are known lead to the formation of massive stars, which evolve quickly
and die as supernovae. The blastwaves from supernova explosions, in addition to
the collisionless shocks at the spiral arms, can perturb the orbits of the
streaming disk matter, occasionally producing rogue exoplanets that can reach
the inner confines of our Solar System. The similarity between the period of
spiral-arm crossings of our Solar System to the period of major extinction
events in the Phanerozoic Eon of the Earth's history, as well as to the period
of the supercontinent cycle (the so-called Wilson Cycle), indicates that the
global environment of the Milky Way Galaxy may have played a major role in
initiating Earth's past tectonic activities.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,208 | Reversing Parallel Programs with Blocks and Procedures | We show how to reverse a while language extended with blocks, local
variables, procedures and the interleaving parallel composition. Annotation is
defined along with a set of operational semantics capable of storing necessary
reversal information, and identifiers are introduced to capture the
interleaving order of an execution. Inversion is defined with a set of
operational semantics that use saved information to undo an execution. We prove
that annotation does not alter the behaviour of the original program, and that
inversion correctly restores the initial program state.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,209 | Detection of the Stellar Intracluster Medium in Perseus (Abell 426) | Hubble Space Telescope photometry from the ACS/WFC and WFPC2 cameras is used
to detect and measure globular clusters (GCs) in the central region of the rich
Perseus cluster of galaxies. A detectable population of Intragalactic GCs is
found extending out to at least 500 kpc from the cluster center. These objects
display luminosity and color (metallicity) distributions that are entirely
normal for GC populations. Extrapolating from the limited spatial coverage of
the HST fields, we estimate very roughly that the entire Perseus cluster should
contain ~50000 or more IGCs, but a targetted wide-field survey will be needed
for a more definitive answer. Separate brief results are presented for the rich
GC systems in NGC 1272 and NGC 1275, the two largest Perseus ellipticals. For
NGC 1272 we find a specific frequency S_N = 8, while for the central giant NGC
1275, S_N ~ 12. In both these giant galaxies, the GC colors are well matched by
bimodal distributions, with the majority in the blue (metal-poor) component.
This preliminary study suggests that Perseus is a prime target for a more
comprehensive deep imaging survey of Intragalactic GCs.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,210 | Binets: fundamental building blocks for phylogenetic networks | Phylogenetic networks are a generalization of evolutionary trees that are
used by biologists to represent the evolution of organisms which have undergone
reticulate evolution. Essentially, a phylogenetic network is a directed acyclic
graph having a unique root in which the leaves are labelled by a given set of
species. Recently, some approaches have been developed to construct
phylogenetic networks from collections of networks on 2- and 3-leaved networks,
which are known as binets and trinets, respectively. Here we study in more
depth properties of collections of binets, one of the simplest possible types
of networks into which a phylogenetic network can be decomposed. More
specifically, we show that if a collection of level-1 binets is compatible with
some binary network, then it is also compatible with a binary level-1 network.
Our proofs are based on useful structural results concerning lowest stable
ancestors in networks. In addition, we show that, although the binets do not
determine the topology of the network, they do determine the number of
reticulations in the network, which is one of its most important parameters. We
also consider algorithmic questions concerning binets. We show that deciding
whether an arbitrary set of binets is compatible with some network is at least
as hard as the well-known Graph Isomorphism problem. However, if we restrict to
level-1 binets, it is possible to decide in polynomial time whether there
exists a binary network that displays all the binets. We also show that to find
a network that displays a maximum number of the binets is NP-hard, but that
there exists a simple polynomial-time 1/3-approximation algorithm for this
problem. It is hoped that these results will eventually assist in the
development of new methods for constructing phylogenetic networks from
collections of smaller networks.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,211 | G-Deformations of maps into projective space | $G$-deformability of maps into projective space is characterised by the
existence of certain Lie algebra valued 1-forms. This characterisation gives a
unified way to obtain well known results regarding deformability in different
geometries.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,212 | Cloudless atmospheres for young low-gravity substellar objects | Atmospheric modeling of low-gravity (VL-G) young brown dwarfs remains a
challenge. The presence of very thick clouds has been suggested because of
their extremely red near-infrared (NIR) spectra, but no cloud models provide a
good fit to the data with a radius compatible with evolutionary models for
these objects. We show that cloudless atmospheres assuming a temperature
gradient reduction caused by fingering convection provides a very good model to
match the observed VL-G NIR spectra. The sequence of extremely red colors in
the NIR for atmospheres with effective temperature from ~2000 K down to ~1200 K
is very well reproduced with predicted radii typical of young low-gravity
objects. Future observations with NIRSPEC and MIRI on the James Webb Space
Telescope (JWST) will provide more constrains in the mid-infrared, helping to
confirm/refute whether or not the NIR reddening is caused by fingering
convection. We suggest that the presence/absence of clouds will be directly
determined by the silicate absorption features that can be observed with MIRI.
JWST will therefore be able to better characterize the atmosphere of these hot
young brown dwarfs and their low-gravity exoplanet analogues.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,213 | On the State of the Art of Evaluation in Neural Language Models | Ongoing innovations in recurrent neural network architectures have provided a
steady influx of apparently state-of-the-art results on language modelling
benchmarks. However, these have been evaluated using differing code bases and
limited computational resources, which represent uncontrolled sources of
experimental variation. We reevaluate several popular architectures and
regularisation methods with large-scale automatic black-box hyperparameter
tuning and arrive at the somewhat surprising conclusion that standard LSTM
architectures, when properly regularised, outperform more recent models. We
establish a new state of the art on the Penn Treebank and Wikitext-2 corpora,
as well as strong baselines on the Hutter Prize dataset.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,214 | Weakly Supervised Audio Source Separation via Spectrum Energy Preserved Wasserstein Learning | Separating audio mixtures into individual instrument tracks has been a long
standing challenging task. We introduce a novel weakly supervised audio source
separation approach based on deep adversarial learning. Specifically, our loss
function adopts the Wasserstein distance which directly measures the
distribution distance between the separated sources and the real sources for
each individual source. Moreover, a global regularization term is added to
fulfill the spectrum energy preservation property regardless separation. Unlike
state-of-the-art weakly supervised models which often involve deliberately
devised constraints or careful model selection, our approach need little prior
model specification on the data, and can be straightforwardly learned in an
end-to-end fashion. We show that the proposed method performs competitively on
public benchmark against state-of-the-art weakly supervised methods.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,215 | Proof of a conjecture of Kløve on permutation codes under the Chebychev distance | Let $d$ be a positive integer and $x$ a real number. Let $A_{d, x}$ be a
$d\times 2d$ matrix with its entries $$ a_{i,j}=\left\{ \begin{array}{ll} x\ \
& \mbox{for} \ 1\leqslant j\leqslant d+1-i, 1\ \ & \mbox{for} \ d+2-i\leqslant
j\leqslant d+i, 0\ \ & \mbox{for} \ d+1+i\leqslant j\leqslant 2d. \end{array}
\right. $$ Further, let $R_d$ be a set of sequences of integers as follows:
$$R_d=\{(\rho_1, \rho_2,\ldots, \rho_d)|1\leqslant \rho_i\leqslant d+i,
1\leqslant i \leqslant d,\ \mbox{and}\ \rho_r\neq \rho_s\ \mbox{for}\ r\neq
s\}.$$ and define $$\Omega_d(x)=\sum_{\rho\in R_d}a_{1,\rho_1}a_{2,
\rho_2}\ldots a_{d,\rho_d}.$$ In order to give a better bound on the size of
spheres of permutation codes under the Chebychev distance, Kl{\o}ve introduced
the above function and conjectured that $$\Omega_d(x)=\sum_{m=0}^d{d\choose
m}(m+1)^d(x-1)^{d-m}.$$ In this paper, we settle down this conjecture
positively.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,216 | Herschel observations of the Galactic HII region RCW 79 | Triggered star formation around HII regions could be an important process.
The Galactic HII region RCW 79 is a prototypical object for triggered high-mass
star formation. We take advantage of Herschel data from the surveys HOBYS,
"Evolution of Interstellar Dust", and Hi-Gal to extract compact sources in this
region, complemented with archival 2MASS, Spitzer, and WISE data to determine
the physical parameters of the sources (e.g., envelope mass, dust temperature,
and luminosity) by fitting the spectral energy distribution. We obtained a
sample of 50 compact sources, 96% of which are situated in the
ionization-compressed layer of cold and dense gas that is characterized by the
column density PDF with a double-peaked lognormal distribution. The 50 sources
have sizes of 0.1-0.4 pc with a typical value of 0.2 pc, temperatures of 11-26
K, envelope masses of 6-760 $M_\odot$, densities of 0.1-44 $\times$ $10^5$
cm$^{-3}$, and luminosities of 19-12712 $L_\odot$. The sources are classified
into 16 class 0, 19 intermediate, and 15 class I objects. Their distribution
follows the evolutionary tracks in the diagram of bolometric luminosity versus
envelope mass (Lbol-Menv) well. A mass threshold of 140 $M_\odot$, determined
from the Lbol-Menv diagram, yields 12 candidate massive dense cores that may
form high-mass stars. The core formation efficiency (CFE) for the 8 massive
condensations shows an increasing trend of the CFE with density. This suggests
that the denser the condensation, the higher the fraction of its mass
transformation into dense cores, as previously observed in other high-mass
star-forming regions.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,217 | Software-Defined Robotics -- Idea & Approach | The methodology of Software-Defined Robotics hierarchical-based and
stand-alone framework can be designed and implemented to program and control
different sets of robots, regardless of their manufacturers' parameters and
specifications, with unified commands and communications. This framework
approach will increase the capability of (re)programming a specific group of
robots during the runtime without affecting the others as desired in the
critical missions and industrial operations, expand the shared bandwidth,
enhance the reusability of code, leverage the computational processing power,
decrease the unnecessary analyses of vast supplemental electrical components
for each robot, as well as get advantages of the most state-of-the-art
industrial trends in the cloud-based computing, Virtual Machines (VM), and
Robot-as-a-Service (RaaS) technologies.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,218 | Stability interchanges in a curved Sitnikov problem | We consider a curved Sitnikov problem, in which an infinitesimal particle
moves on a circle under the gravitational influence of two equal masses in
Keplerian motion within a plane perpendicular to that circle. There are two
equilibrium points, whose stability we are studying. We show that one of the
equilibrium points undergoes stability interchanges as the semi-major axis of
the Keplerian ellipses approaches the diameter of that circle. To derive this
result, we first formulate and prove a general theorem on stability
interchanges, and then we apply it to our model. The motivation for our model
resides with the $n$-body problem in spaces of constant curvature.
| 0 | 1 | 1 | 0 | 0 | 0 |
20,219 | Hardy Spaces over Half-strip Domains | We define Hardy spaces $H^p(\Omega_\pm)$ on half-strip domain~$\Omega_+$ and
$\Omega_-= \mathbb{C}\setminus\overline{\Omega_+}$, where $0<p<\infty$, and
prove that functions in $H^p(\Omega_\pm)$ has non-tangential boundary limit
a.e. on $\Gamma$, the common boundary of $\Omega_\pm$. We then prove that
Cauchy integral of functions in $L^p(\Gamma)$ are in $H^p(\Omega_\pm)$, where
$1<p<\infty$, that is, Cauchy transform is bounded. Besides, if $1\leqslant
p<\infty$, then $H^p(\Omega_\pm)$ functions are the Cauchy integral of their
non-tangential boundary limits. We also establish an isomorphism between
$H^p(\Omega_\pm)$ and $H^p(\mathbb{C}_\pm)$, the classical Hardy spaces over
upper and lower half complex planes.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,220 | An Optimization Framework with Flexible Inexact Inner Iterations for Nonconvex and Nonsmooth Programming | In recent years, numerous vision and learning tasks have been (re)formulated
as nonconvex and nonsmooth programmings(NNPs). Although some algorithms have
been proposed for particular problems, designing fast and flexible optimization
schemes with theoretical guarantee is a challenging task for general NNPs. It
has been investigated that performing inexact inner iterations often benefit to
special applications case by case, but their convergence behaviors are still
unclear. Motivated by these practical experiences, this paper designs a novel
algorithmic framework, named inexact proximal alternating direction method
(IPAD) for solving general NNPs. We demonstrate that any numerical algorithms
can be incorporated into IPAD for solving subproblems and the convergence of
the resulting hybrid schemes can be consistently guaranteed by a series of
simple error conditions. Beyond the guarantee in theory, numerical experiments
on both synthesized and real-world data further demonstrate the superiority and
flexibility of our IPAD framework for practical use.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,221 | Attitude and angular velocity tracking for a rigid body using geometric methods on the two-sphere | The control task of tracking a reference pointing direction (the attitude
about the pointing direction is irrelevant) while obtaining a desired angular
velocity (PDAV) around the pointing direction using geometric techniques is
addressed here. Existing geometric controllers developed on the two-sphere only
address the tracking of a reference pointing direction while driving the
angular velocity about the pointing direction to zero. In this paper a tracking
controller on the two-sphere, able to address the PDAV control task, is
developed globally in a geometric frame work, to avoid problems related to
other attitude representations such as unwinding (quaternions) or singularities
(Euler angles). An attitude error function is constructed resulting in a
control system with desired tracking performance for rotational maneuvers with
large initial attitude/angular velocity errors and the ability to negotiate
bounded modeling inaccuracies. The tracking ability of the developed control
system is evaluated by comparing its performance with an existing geometric
controller on the two-sphere and by numerical simulations, showing improved
performance for large initial attitude errors, smooth transitions between
desired angular velocities and the ability to negotiate bounded modeling
inaccuracies.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,222 | Miscomputation in software: Learning to live with errors | Computer programs do not always work as expected. In fact, ominous warnings
about the desperate state of the software industry continue to be released with
almost ritualistic regularity. In this paper, we look at the 60 years history
of programming and at the different practical methods that software community
developed to live with programming errors. We do so by observing a class of
students discussing different approaches to programming errors. While learning
about the different methods for dealing with errors, we uncover basic
assumptions that proponents of different paradigms follow. We learn about the
mathematical attempt to eliminate errors through formal methods, scientific
method based on testing, a way of building reliable systems through engineering
methods, as well as an artistic approach to live coding that accepts errors as
a creative inspiration. This way, we can explore the differences and
similarities among the different paradigms. By inviting proponents of different
methods into a single discussion, we hope to open potential for new thinking
about errors. When should we use which of the approaches? And what can software
development learn from mathematics, science, engineering and art? When
programming or studying programming, we are often enclosed in small communities
and we take our basic assumptions for granted. Through the discussion in this
paper, we attempt to map the large and rich space of programming ideas and
provide reference points for exploring, perhaps foreign, ideas that can
challenge some of our assumptions.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,223 | The Combinatorics of Weighted Vector Compositions | A vector composition of a vector $\mathbf{\ell}$ is a matrix $\mathbf{A}$
whose rows sum to $\mathbf{\ell}$. We define a weighted vector composition as a
vector composition in which the column values of $\mathbf{A}$ may appear in
different colors. We study vector compositions from different viewpoints: (1)
We show how they are related to sums of random vectors and (2) how they allow
to derive formulas for partial derivatives of composite functions. (3) We study
congruence properties of the number of weighted vector compositions, for fixed
and arbitrary number of parts, many of which are analogous to those of ordinary
binomial coefficients and related quantities. Via the Central Limit Theorem and
their multivariate generating functions, (4) we also investigate the asymptotic
behavior of several special cases of numbers of weighted vector compositions.
Finally, (5) we conjecture an extension of a primality criterion due to Mann
and Shanks in the context of weighted vector compositions.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,224 | Fast Linear Model for Knowledge Graph Embeddings | This paper shows that a simple baseline based on a Bag-of-Words (BoW)
representation learns surprisingly good knowledge graph embeddings. By casting
knowledge base completion and question answering as supervised classification
problems, we observe that modeling co-occurences of entities and relations
leads to state-of-the-art performance with a training time of a few minutes
using the open sourced library fastText.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,225 | Deep learning for plasma tomography using the bolometer system at JET | Deep learning is having a profound impact in many fields, especially those
that involve some form of image processing. Deep neural networks excel in
turning an input image into a set of high-level features. On the other hand,
tomography deals with the inverse problem of recreating an image from a number
of projections. In plasma diagnostics, tomography aims at reconstructing the
cross-section of the plasma from radiation measurements. This reconstruction
can be computed with neural networks. However, previous attempts have focused
on learning a parametric model of the plasma profile. In this work, we use a
deep neural network to produce a full, pixel-by-pixel reconstruction of the
plasma profile. For this purpose, we use the overview bolometer system at JET,
and we introduce an up-convolutional network that has been trained and tested
on a large set of sample tomograms. We show that this network is able to
reproduce existing reconstructions with a high level of accuracy, as measured
by several metrics.
| 0 | 1 | 0 | 1 | 0 | 0 |
20,226 | Assembly Bias and Splashback in Galaxy Clusters | We use publicly available data for the Millennium Simulation to explore the
implications of the recent detection of assembly bias and splashback signatures
in a large sample of galaxy clusters. These were identified in the SDSS/DR8
photometric data by the redMaPPer algorithm and split into high- and
low-concentration subsamples based on the projected positions of cluster
members. We use simplified versions of these procedures to build cluster
samples of similar size from the simulation data. These match the observed
samples quite well and show similar assembly bias and splashback signals.
Previous theoretical work has found the logarithmic slope of halo density
profiles to have a well-defined minimum whose depth decreases and whose radius
increases with halo concentration. Projected profiles for the observed and
simulated cluster samples show trends with concentration which are opposite to
these predictions. In addition, for high-concentration clusters the minimum
slope occurs at significantly smaller radius than predicted. We show that these
discrepancies all reflect confusion between splashback features and features
imposed on the profiles by the cluster identification and concentration
estimation procedures. The strong apparent assembly bias is not reflected in
the three-dimensional distribution of matter around clusters. Rather it is a
consequence of the preferential contamination of low-concentration clusters by
foreground or background groups.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,227 | Accelerating Kernel Classifiers Through Borders Mapping | Support vector machines (SVM) and other kernel techniques represent a family
of powerful statistical classification methods with high accuracy and broad
applicability. Because they use all or a significant portion of the training
data, however, they can be slow, especially for large problems. Piecewise
linear classifiers are similarly versatile, yet have the additional advantages
of simplicity, ease of interpretation and, if the number of component linear
classifiers is not too large, speed. Here we show how a simple, piecewise
linear classifier can be trained from a kernel-based classifier in order to
improve the classification speed. The method works by finding the root of the
difference in conditional probabilities between pairs of opposite classes to
build up a representation of the decision boundary. When tested on 17 different
datasets, it succeeded in improving the classification speed of a SVM for 9 of
them by factors as high as 88 times or more. The method is best suited to
problems with continuum features data and smooth probability functions. Because
the component linear classifiers are built up individually from an existing
classifier, rather than through a simultaneous optimization procedure, the
classifier is also fast to train.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,228 | Information Criterion for Minimum Cross-Entropy Model Selection | This paper considers the problem of approximating a density when it can be
evaluated up to a normalizing constant at a finite number of points. This
density approximation problem is ubiquitous in machine learning, such as
approximating a posterior density for Bayesian inference and estimating an
optimal density for importance sampling. Approximating the density with a
parametric model can be cast as a model selection problem. This problem cannot
be addressed with traditional approaches that maximize the (marginal)
likelihood of a model, for example, using the Akaike information criterion
(AIC) or Bayesian information criterion (BIC). We instead aim to minimize the
cross-entropy that gauges the deviation of a parametric model from the target
density. We propose a novel information criterion called the cross-entropy
information criterion (CIC) and prove that the CIC is an asymptotically
unbiased estimator of the cross-entropy (up to a multiplicative constant) under
some regularity conditions. We propose an iterative method to approximate the
target density by minimizing the CIC. We demonstrate that the proposed method
selects a parametric model that well approximates the target density.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,229 | Scaling relations in large-Prandtl-number natural thermal convection | In this study we follow Grossmann and Lohse, Phys. Rev. Lett. 86 (2001), who
derived various scalings regimes for the dependence of the Nusselt number $Nu$
and the Reynolds number $Re$ on the Rayleigh number $Ra$ and the Prandtl number
$Pr$. We focus on theoretical arguments as well as on numerical simulations for
the case of large-$Pr$ natural thermal convection. Based on an analysis of
self-similarity of the boundary layer equations, we derive that in this case
the limiting large-$Pr$ boundary-layer dominated regime is I$_\infty^<$,
introduced and defined in [1], with the scaling relations $Nu\sim
Pr^0\,Ra^{1/3}$ and $Re\sim Pr^{-1}\,Ra^{2/3}$. Our direct numerical
simulations for $Ra$ from $10^4$ to $10^9$ and $Pr$ from 0.1 to 200 show that
the regime I$_\infty^<$ is almost indistinguishable from the regime
III$_\infty$, where the kinetic dissipation is bulk-dominated. With increasing
$Ra$, the scaling relations undergo a transition to those in IV$_u$ of
reference [1], where the thermal dissipation is determined by its bulk
contribution.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,230 | The renormalization method from continuous to discrete dynamical systems: asymptotic solutions, reductions and invariant manifolds | The renormalization method based on the Taylor expansion for asymptotic
analysis of differential equations is generalized to difference equations. The
proposed renormalization method is based on the Newton-Maclaurin expansion.
Several basic theorems on the renormalization method are proven. Some
interesting applications are given, including asymptotic solutions of quantum
anharmonic oscillator and discrete boundary layer, the reductions and invariant
manifolds of some discrete dynamics systems. Furthermore, the homotopy
renormalization method based on the Newton-Maclaurin expansion is proposed and
applied to those difference equations including no a small parameter.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,231 | A Stress/Displacement Virtual Element Method for Plane Elasticity Problems | The numerical approximation of 2D elasticity problems is considered, in the
framework of the small strain theory and in connection with the mixed
Hellinger-Reissner variational formulation. A low-order Virtual Element Method
(VEM) with a-priori symmetric stresses is proposed. Several numerical tests are
provided, along with a rigorous stability and convergence analysis.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,232 | Estimation of the lead-lag parameter between two stochastic processes driven by fractional Brownian motions | In this paper, we consider the problem of estimating the lead-lag parameter
between two stochastic processes driven by fractional Brownian motions (fBMs)
of the Hurst parameter greater than 1/2. First we propose a lead-lag model
between two stochastic processes involving fBMs, and then construct a
consistent estimator of the lead-lag parameter with possible convergence rate.
Our estimator has the following two features. Firstly, we can construct the
lead-lag estimator without using the Hurst parameters of the underlying fBMs.
Secondly, our estimator can deal with some non-synchronous and irregular
observations. We explicitly calculate possible convergence rate when the
observation times are (1) synchronous and equidistant, and (2) given by the
Poisson sampling scheme. We also present numerical simulations of our results
using the R package YUIMA.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,233 | Multi-resolution polymer Brownian dynamics with hydrodynamic interactions | A polymer model given in terms of beads, interacting through Hookean springs
and hydrodynamic forces, is studied. Brownian dynamics description of this
bead-spring polymer model is extended to multiple resolutions. Using this
multiscale approach, a modeller can efficiently look at different regions of
the polymer in different spatial and temporal resolutions with scalings given
for the number of beads, statistical segment length and bead radius in order to
maintain macro-scale properties of the polymer filament. The Boltzmann
distribution of a Gaussian chain for differing statistical segment lengths
gives a Langevin equation for the multi-resolution model with a mobility tensor
for different bead sizes. Using the pre-averaging approximation, the
translational diffusion coefficient is obtained as a function of the inverse of
a matrix and then in closed form in the long-chain limit. This is then
confirmed with numerical experiments.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,234 | Learning Theory of Distributed Regression with Bias Corrected Regularization Kernel Network | Distributed learning is an effective way to analyze big data. In distributed
regression, a typical approach is to divide the big data into multiple blocks,
apply a base regression algorithm on each of them, and then simply average the
output functions learnt from these blocks. Since the average process will
decrease the variance, not the bias, bias correction is expected to improve the
learning performance if the base regression algorithm is a biased one.
Regularization kernel network is an effective and widely used method for
nonlinear regression analysis. In this paper we will investigate a bias
corrected version of regularization kernel network. We derive the error bounds
when it is applied to a single data set and when it is applied as a base
algorithm in distributed regression. We show that, under certain appropriate
conditions, the optimal learning rates can be reached in both situations.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,235 | Cluster-based Kriging Approximation Algorithms for Complexity Reduction | Kriging or Gaussian Process Regression is applied in many fields as a
non-linear regression model as well as a surrogate model in the field of
evolutionary computation. However, the computational and space complexity of
Kriging, that is cubic and quadratic in the number of data points respectively,
becomes a major bottleneck with more and more data available nowadays. In this
paper, we propose a general methodology for the complexity reduction, called
cluster Kriging, where the whole data set is partitioned into smaller clusters
and multiple Kriging models are built on top of them. In addition, four Kriging
approximation algorithms are proposed as candidate algorithms within the new
framework. Each of these algorithms can be applied to much larger data sets
while maintaining the advantages and power of Kriging. The proposed algorithms
are explained in detail and compared empirically against a broad set of
existing state-of-the-art Kriging approximation methods on a well-defined
testing framework. According to the empirical study, the proposed algorithms
consistently outperform the existing algorithms. Moreover, some practical
suggestions are provided for using the proposed algorithms.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,236 | How to model fake news | Over the past three years it has become evident that fake news is a danger to
democracy. However, until now there has been no clear understanding of how to
define fake news, much less how to model it. This paper addresses both these
issues. A definition of fake news is given, and two approaches for the
modelling of fake news and its impact in elections and referendums are
introduced. The first approach, based on the idea of a representative voter, is
shown to be suitable to obtain a qualitative understanding of phenomena
associated with fake news at a macroscopic level. The second approach, based on
the idea of an election microstructure, describes the collective behaviour of
the electorate by modelling the preferences of individual voters. It is shown
through a simulation study that the mere knowledge that pieces of fake news may
be in circulation goes a long way towards mitigating the impact of fake news.
| 1 | 0 | 0 | 0 | 0 | 1 |
20,237 | Analysis of Dropout in Online Learning | Deep learning is the state-of-the-art in fields such as visual object
recognition and speech recognition. This learning uses a large number of layers
and a huge number of units and connections. Therefore, overfitting is a serious
problem with it, and the dropout which is a kind of regularization tool is
used. However, in online learning, the effect of dropout is not well known.
This paper presents our investigation on the effect of dropout in online
learning. We analyzed the effect of dropout on convergence speed near the
singular point. Our results indicated that dropout is effective in online
learning. Dropout tends to avoid the singular point for convergence speed near
that point.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,238 | TADPOLE Challenge: Prediction of Longitudinal Evolution in Alzheimer's Disease | The Alzheimer's Disease Prediction Of Longitudinal Evolution (TADPOLE)
Challenge compares the performance of algorithms at predicting future evolution
of individuals at risk of Alzheimer's disease. TADPOLE Challenge participants
train their models and algorithms on historical data from the Alzheimer's
Disease Neuroimaging Initiative (ADNI) study or any other datasets to which
they have access. Participants are then required to make monthly forecasts over
a period of 5 years from January 2018, of three key outcomes for ADNI-3
rollover participants: clinical diagnosis, Alzheimer's Disease Assessment Scale
Cognitive Subdomain (ADAS-Cog13), and total volume of the ventricles. These
individual forecasts are later compared with the corresponding future
measurements in ADNI-3 (obtained after the TADPOLE submission deadline). The
first submission phase of TADPOLE was open for prize-eligible submissions
between 15 June and 15 November 2017. The submission system remains open via
the website: this https URL, although since 15 November
2017 submissions are not eligible for the first round of prizes. This paper
describes the design of the TADPOLE Challenge.
| 0 | 0 | 0 | 1 | 1 | 0 |
20,239 | A Deep Cascade of Convolutional Neural Networks for MR Image Reconstruction | The acquisition of Magnetic Resonance Imaging (MRI) is inherently slow.
Inspired by recent advances in deep learning, we propose a framework for
reconstructing MR images from undersampled data using a deep cascade of
convolutional neural networks to accelerate the data acquisition process. We
show that for Cartesian undersampling of 2D cardiac MR images, the proposed
method outperforms the state-of-the-art compressed sensing approaches, such as
dictionary learning-based MRI (DLMRI) reconstruction, in terms of
reconstruction error, perceptual quality and reconstruction speed for both
3-fold and 6-fold undersampling. Compared to DLMRI, the error produced by the
method proposed is approximately twice as small, allowing to preserve
anatomical structures more faithfully. Using our method, each image can be
reconstructed in 23 ms, which is fast enough to enable real-time applications.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,240 | Advances in Joint CTC-Attention based End-to-End Speech Recognition with a Deep CNN Encoder and RNN-LM | We present a state-of-the-art end-to-end Automatic Speech Recognition (ASR)
model. We learn to listen and write characters with a joint Connectionist
Temporal Classification (CTC) and attention-based encoder-decoder network. The
encoder is a deep Convolutional Neural Network (CNN) based on the VGG network.
The CTC network sits on top of the encoder and is jointly trained with the
attention-based decoder. During the beam search process, we combine the CTC
predictions, the attention-based decoder predictions and a separately trained
LSTM language model. We achieve a 5-10\% error reduction compared to prior
systems on spontaneous Japanese and Chinese speech, and our end-to-end model
beats out traditional hybrid ASR systems.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,241 | The Social and Work Structure of an Afterschool Math Club | This study focuses on the social structure and interpersonal dynamics of an
afterschool math club for middle schoolers. Using social network analysis, two
networks were formed and analyzed: The network of friendship relationships and
the network of working relationships. The interconnections and correlations
between friendship relationships, working relationships, and student opinion
surveys are studied. A core working group of talented students emerged from the
network of working relations. This group acted a central go-between for other
members in the club. This core working group expanded into largest friendship
group in the friendship network. Although there were working isolates, they
were not found to be socially isolated. Students who were less popular tended
to report a greater favorable impact from club participation. Implications for
the study of the social structure of afterschool STEM clubs and classrooms are
discussed.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,242 | Fourier Transform of Schwartz Algebras on Groups in the Harish-Chandra class | It is well-known that the Harish-Chandra transform, $f\mapsto\mathcal{H}f,$
is a topological isomorphism of the spherical (Schwartz) convolution algebra
$\mathcal{C}^{p}(G//K)$ (where $K$ is a maximal compact subgroup of any
arbitrarily chosen group $G$ in the Harish-Chandra class and $0<p\leq2$) onto
the (Schwartz) multiplication algebra
$\bar{\mathcal{Z}}({\mathfrak{F}}^{\epsilon})$ (of $\mathfrak{w}-$invariant
members of $\mathcal{Z}({\mathfrak{F}}^{\epsilon}),$ with $\epsilon=(2/p)-1$).
The same cannot however be said of the full Schwartz convolution algebra
$\mathcal{C}^{p}(G),$ except for few specific examples of groups (notably
$G=SL(2,\mathbb{R})$) and for some notable values of $p$ (with restrictions on
$G$ and/or on $\mathcal{C}^{p}(G)$). Nevertheless the full Harish-Chandra
Plancherel formula on $G$ is known for all of
$\mathcal{C}^{2}(G)=:\mathcal{C}(G).$ In order to then understand the structure
of Harish-Chandra transform more clearly and to compute the image of
$\mathcal{C}^{p}(G)$ under it (without any restriction) we derive an absolutely
convergent series expansion (in terms of known functions) for the
Harish-Chandra transform by an application of the full Plancherel formula on
$G.$ This leads to a computation of the image of $\mathcal{C}(G)$ under the
Harish-Chandra transform which may be seen as a concrete realization of
Arthur's result and be easily extended to all of $\mathcal{C}^{p}(G)$ in much
the same way as it is known in the work of Trombi and Varadarajan.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,243 | Mutual Information and Optimality of Approximate Message-Passing in Random Linear Estimation | We consider the estimation of a signal from the knowledge of its noisy linear
random Gaussian projections. A few examples where this problem is relevant are
compressed sensing, sparse superposition codes, and code division multiple
access. There has been a number of works considering the mutual information for
this problem using the replica method from statistical physics. Here we put
these considerations on a firm rigorous basis. First, we show, using a
Guerra-Toninelli type interpolation, that the replica formula yields an upper
bound to the exact mutual information. Secondly, for many relevant practical
cases, we present a converse lower bound via a method that uses spatial
coupling, state evolution analysis and the I-MMSE theorem. This yields a single
letter formula for the mutual information and the minimal-mean-square error for
random Gaussian linear estimation of all discrete bounded signals. In addition,
we prove that the low complexity approximate message-passing algorithm is
optimal outside of the so-called hard phase, in the sense that it
asymptotically reaches the minimal-mean-square error. In this work spatial
coupling is used primarily as a proof technique. However our results also prove
two important features of spatially coupled noisy linear random Gaussian
estimation. First there is no algorithmically hard phase. This means that for
such systems approximate message-passing always reaches the minimal-mean-square
error. Secondly, in a proper limit the mutual information associated to such
systems is the same as the one of uncoupled linear random Gaussian estimation.
| 1 | 1 | 1 | 0 | 0 | 0 |
20,244 | A Memristor-Based Optimization Framework for AI Applications | Memristors have recently received significant attention as ubiquitous
device-level components for building a novel generation of computing systems.
These devices have many promising features, such as non-volatility, low power
consumption, high density, and excellent scalability. The ability to control
and modify biasing voltages at the two terminals of memristors make them
promising candidates to perform matrix-vector multiplications and solve systems
of linear equations. In this article, we discuss how networks of memristors
arranged in crossbar arrays can be used for efficiently solving optimization
and machine learning problems. We introduce a new memristor-based optimization
framework that combines the computational merit of memristor crossbars with the
advantages of an operator splitting method, alternating direction method of
multipliers (ADMM). Here, ADMM helps in splitting a complex optimization
problem into subproblems that involve the solution of systems of linear
equations. The capability of this framework is shown by applying it to linear
programming, quadratic programming, and sparse optimization. In addition to
ADMM, implementation of a customized power iteration (PI) method for
eigenvalue/eigenvector computation using memristor crossbars is discussed. The
memristor-based PI method can further be applied to principal component
analysis (PCA). The use of memristor crossbars yields a significant speed-up in
computation, and thus, we believe, has the potential to advance optimization
and machine learning research in artificial intelligence (AI).
| 1 | 0 | 0 | 1 | 0 | 0 |
20,245 | Estimating parameters of a directed weighted graph model with beta-distributed edge-weights | We introduce a directed, weighted random graph model, where the edge-weights
are independent and beta-distributed with parameters depending on their
endpoints. We will show that the row- and column-sums of the transformed
edge-weight matrix are sufficient statistics for the parameters, and use the
theory of exponential families to prove that the ML estimate of the parameters
exists and is unique. Then an algorithm to find this estimate is introduced
together with convergence proof that uses properties of the digamma function.
Simulation results and applications are also presented.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,246 | On the Performance of Reduced-Complexity Transmit/Receive Diversity Systems over MIMO-V2V Channel Model | In this letter, we investigate the performance of multiple-input
multiple-output techniques in a vehicle-to-vehicle communication system. We
consider both transmit antenna selection with maximal-ratio combining and
transmit antenna selection with selection combining. The channel propagation
model between two vehicles is represented as n*Rayleigh distribution, which has
been shown to be a realistic model for vehicle-to-vehicle communication
scenarios. We derive tight analytical expressions for the outage probability
and amount of fading of the post-processing signal-to-noise ratio.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,247 | Internal migration and education: A cross-national comparison | Migration the main process shaping patterns of human settlement within and
between countries. It is widely acknowledged to be integral to the process of
human development as it plays a significant role in enhancing educational
outcomes. At regional and national levels, internal migration underpins the
efficient functioning of the economy by bringing knowledge and skills to the
locations where they are needed. It is the multi-dimensional nature of
migration that underlines its significance in the process of human development.
Human mobility extends in the spatial domain from local travel to international
migration, and in the temporal dimension from short-term stays to permanent
relocations. Classification and measurement of such phenomena is inevitably
complex, which has severely hindered progress in comparative research, with
very few large-scale cross-national comparisons of migration. The linkages
between migration and education have been explored in a separate line of
inquiry that has predominantly focused on country-specific analyses as to the
ways in which migration affects educational outcomes and how educational
attainment affects migration behaviour. A recurrent theme has been the
educational selectivity of migrants, which in turn leads to an increase of
human capital in some regions, primarily cities, at the expense of others.
Questions have long been raised as to the links between education and migration
in response to educational expansion, but have not yet been fully answered
because of the absence, until recently, of adequate data for comparative
analysis of migration. In this paper, we bring these two separate strands of
research together to systematically explore links between internal migration
and education across a global sample of 57 countries at various stages of
development, using data drawn from the IPUMS database.
| 0 | 0 | 0 | 0 | 0 | 1 |
20,248 | Free differential Lie Rota-Baxter algebras and Gröbner-Shirshov bases | We establish the Gröbner-Shirshov bases theory for differential Lie
$\Omega$-algebras. As an application, we give a linear basis of a free
differential Lie Rota-Baxter algebra on a set.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,249 | Conversion Rate Optimization through Evolutionary Computation | Conversion optimization means designing a web interface so that as many users
as possible take a desired action on it, such as register or purchase. Such
design is usually done by hand, testing one change at a time through A/B
testing, or a limited number of combinations through multivariate testing,
making it possible to evaluate only a small fraction of designs in a vast
design space. This paper describes Sentient Ascend, an automatic conversion
optimization system that uses evolutionary optimization to create effective web
interface designs. Ascend makes it possible to discover and utilize
interactions between the design elements that are difficult to identify
otherwise. Moreover, evaluation of design candidates is done in parallel
online, i.e. with a large number of real users interacting with the system. A
case study on an existing media site shows that significant improvements (i.e.
over 43%) are possible beyond human design. Ascend can therefore be seen as an
approach to massively multivariate conversion optimization, based on a
massively parallel interactive evolution.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,250 | Emergence of spatial curvature | This paper investigates the phenomenon of emergence of spatial curvature.
This phenomenon is absent in the Standard Cosmological Model, which has a flat
and fixed spatial curvature (small perturbations are considered in the Standard
Cosmological Model but their global average vanishes, leading to spatial
flatness at all times). This paper shows that with the nonlinear growth of
cosmic structures the global average deviates from zero. The analysis is based
on the {\em silent universes} (a wide class of inhomogeneous cosmological
solutions of the Einstein equations). The initial conditions are set in the
early universe as perturbations around the $\Lambda$CDM model with $\Omega_m =
0.31$, $\Omega_\Lambda = 0.69$, and $H_0 = 67.8$ km s$^{-1}$ Mpc$^{-1}$. As the
growth of structures becomes nonlinear, the model deviates from the
$\Lambda$CDM model, and at the present instant if averaged over a domain ${\cal
D}$ with volume $V = (2150\,{\rm Mpc})^3$ (at these scales the cosmic variance
is negligibly small) gives: $\Omega_m^{\cal D} = 0.22$, $\Omega_\Lambda^{\cal
D} = 0.61$, $\Omega_{\cal R}^{\cal D} = 0.15$ (in the FLRW limit $\Omega_{\cal
R}^{\cal D} \to \Omega_k$), and $\langle H \rangle_{\cal D} = 72.2$ km s$^{-1}$
Mpc$^{-1}$. Given the fact that low-redshift observations favor higher values
of the Hubble constant and lower values of matter density, compared to the CMB
constraints, the emergence of the spatial curvature in the low-redshift
universe could be a possible solution to these discrepancies.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,251 | Asymptotic generalized bivariate extreme with random index | In many biological, agricultural, military activity problems and in some
quality control problems, it is almost impossible to have a fixed sample size,
because some observations are always lost for various reasons. Therefore, the
sample size itself is considered frequently to be a random variable (rv). The
class of limit distribution functions (df's) of the random bivariate extreme
generalized order statistics (GOS) from independent and identically distributed
RV's are fully characterized. When the random sample size is assumed to be
independent of the basic variables and its df is assumed to converge weakly to
a non-degenerate limit, the necessary and sufficient conditions for the weak
convergence of the random bivariate extreme GOS are obtained. Furthermore, when
the interrelation of the random size and the basic rv's is not restricted,
sufficient conditions for the convergence and the forms of the limit df's are
deduced. Illustrative examples are given which lend further support to our
theoretical results.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,252 | On problems in the calculus of variations in increasingly elongated domains | We consider minimization problems in the calculus of variations set in a
sequence of domains the size of which tends to infinity in certain directions
and such that the data only depend on the coordinates in the directions that
remain constant. We study the asymptotic behavior of minimizers in various
situations and show that they converge in an appropriate sense toward
minimizers of a related energy functional in the constant directions.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,253 | Inequalities related to Symmetrized Harmonic Convex Functions | In this paper, we extend the Hermite-Hadamard type $\dot{I}$scan inequality
to the class of symmetrized harmonic convex functions. The corresponding
version for harmonic h-convex functions is also investigated. Furthermore, we
establish Hermite-Hadamard type inequalites for the product of a harmonic
convex function with a symmetrized harmonic convex function.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,254 | Parameter estimation for fractional Ornstein-Uhlenbeck processes of general Hurst parameter | This paper provides several statistical estimators for the drift and
volatility parameters of an Ornstein-Uhlenbeck process driven by fractional
Brownian motion, whose observations can be made either continuously or at
discrete time instants. First and higher order power variations are used to
estimate the volatility parameter. The almost sure convergence of the
estimators and the corresponding central limit theorems are obtained for all
the Hurst parameter range $H\in (0, 1)$. The least squares estimator is used
for the drift parameter. A central limit theorem is proved when the Hurst
parameter $H \in (0, 1/2)$ and a noncentral limit theorem is proved for
$H\in[3/4, 1)$. Thus, the open problem left in the paper by Hu and Nualart
(2010) is completely solved, where a central limit theorem for least squares
estimator is proved for $H\in [1/2, 3/4)$.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,255 | E-polynomials of $PGL(2,\mathbb{C})$-character varieties of surface groups | In this paper, we compute the E-polynomials of the
$PGL(2,\mathbb{C})$-character varieties associated to surfaces of genus $g$
with one puncture, for any holonomy around it, and compare it with its
Langlands dual case, $SL(2,\mathbb{C})$. The study is based on the
stratification of the space of representations and on the analysis of the
behaviour of the E-polynomial under fibrations.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,256 | Security Analysis of Cache Replacement Policies | Modern computer architectures share physical resources between different
programs in order to increase area-, energy-, and cost-efficiency.
Unfortunately, sharing often gives rise to side channels that can be exploited
for extracting or transmitting sensitive information. We currently lack
techniques for systematic reasoning about this interplay between security and
efficiency. In particular, there is no established way for quantifying security
properties of shared caches.
In this paper, we propose a novel model that enables us to characterize
important security properties of caches. Our model encompasses two aspects: (1)
The amount of information that can be absorbed by a cache, and (2) the amount
of information that can effectively be extracted from the cache by an
adversary. We use our model to compute both quantities for common cache
replacement policies (FIFO, LRU, and PLRU) and to compare their isolation
properties. We further show how our model for information extraction leads to
an algorithm that can be used to improve the bounds delivered by the CacheAudit
static analyzer.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,257 | Supercongruences related to ${}_3F_2(1)$ involving harmonic numbers | We show various supercongruences for truncated series which involve central
binomial coefficients and harmonic numbers. The corresponding infinite series
are also evaluated.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,258 | Dynamic Layer Normalization for Adaptive Neural Acoustic Modeling in Speech Recognition | Layer normalization is a recently introduced technique for normalizing the
activities of neurons in deep neural networks to improve the training speed and
stability. In this paper, we introduce a new layer normalization technique
called Dynamic Layer Normalization (DLN) for adaptive neural acoustic modeling
in speech recognition. By dynamically generating the scaling and shifting
parameters in layer normalization, DLN adapts neural acoustic models to the
acoustic variability arising from various factors such as speakers, channel
noises, and environments. Unlike other adaptive acoustic models, our proposed
approach does not require additional adaptation data or speaker information
such as i-vectors. Moreover, the model size is fixed as it dynamically
generates adaptation parameters. We apply our proposed DLN to deep
bidirectional LSTM acoustic models and evaluate them on two benchmark datasets
for large vocabulary ASR experiments: WSJ and TED-LIUM release 2. The
experimental results show that our DLN improves neural acoustic models in terms
of transcription accuracy by dynamically adapting to various speakers and
environments.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,259 | First-Order vs. Second-Order Encodings for LTLf-to-Automata Translation | Translating formulas of Linear Temporal Logic (LTL) over finite traces, or
LTLf, to symbolic Deterministic Finite Automata (DFA) plays an important role
not only in LTLf synthesis, but also in synthesis for Safety LTL formulas. The
translation is enabled by using MONA, a powerful tool for symbolic, BDD-based,
DFA construction from logic specifications. Recent works used a first-order
encoding of LTLf formulas to translate LTLf to First Order Logic (FOL), which
is then fed to MONA to get the symbolic DFA. This encoding was shown to perform
well, but other encodings have not been studied. Specifically, the natural
question of whether second-order encoding, which has significantly simpler
quantificational structure, can outperform first-order encoding remained open.
In this paper we address this challenge and study second-order encodings for
LTLf formulas. We first introduce a specific MSO encoding that captures the
semantics of LTLf in a natural way and prove its correctness. We then explore
is a Compact MSO encoding, which benefits from automata-theoretic minimization,
thus suggesting a possible practical advantage. To that end, we propose a
formalization of symbolic DFA in second-order logic, thus developing a novel
connection between BDDs and MSO. We then show by empirical evaluations that the
first-order encoding does perform better than both second-order encodings. The
conclusion is that first-order encoding is a better choice than second-order
encoding in LTLf-to-Automata translation.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,260 | Regularity results and parametrices of semi-linear boundary problems of product type | This short note describes the benefit one obtains from a specific
construction of a family of parametrices for a class of elliptic boundary value
problems perturbed by non-linear terms of product type. The construction is
based on the Boutet de Monvel calculus of pseudo-differential boundary
operators for the linear elliptic parts, and on paradifferential operators for
the product terms.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,261 | $\left( β, \varpi \right)$-stability for cross-validation and the choice of the number of folds | In this paper, we introduce a new concept of stability for cross-validation,
called the $\left( \beta, \varpi \right)$-stability, and use it as a new
perspective to build the general theory for cross-validation. The $\left(
\beta, \varpi \right)$-stability mathematically connects the generalization
ability and the stability of the cross-validated model via the Rademacher
complexity. Our result reveals mathematically the effect of cross-validation
from two sides: on one hand, cross-validation picks the model with the best
empirical generalization ability by validating all the alternatives on test
sets; on the other hand, cross-validation may compromise the stability of the
model selection by causing subsampling error. Moreover, the difference between
training and test errors in q\textsuperscript{th} round, sometimes referred to
as the generalization error, might be autocorrelated on q. Guided by the ideas
above, the $\left( \beta, \varpi \right)$-stability help us derivd a new class
of Rademacher bounds, referred to as the one-round/convoluted Rademacher
bounds, for the stability of cross-validation in both the i.i.d.\ and
non-i.i.d.\ cases. For both light-tail and heavy-tail losses, the new bounds
quantify the stability of the one-round/average test error of the
cross-validated model in terms of its one-round/average training error, the
sample sizes $n$, number of folds $K$, the tail property of the loss (encoded
as Orlicz-$\Psi_\nu$ norms) and the Rademacher complexity of the model class
$\Lambda$. The new class of bounds not only quantitatively reveals the
stability of the generalization ability of the cross-validated model, it also
shows empirically the optimal choice for number of folds $K$, at which the
upper bound of the one-round/average test error is lowest, or, to put it in
another way, where the test error is most stable.
| 1 | 0 | 1 | 1 | 0 | 0 |
20,262 | Directional convexity of harmonic mappings | The convolution properties are discussed for the complex-valued harmonic
functions in the unit disk $\mathbb{D}$ constructed from the harmonic shearing
of the analytic function $\phi(z):=\int_0^z
(1/(1-2\xi\textit{e}^{\textit{i}\mu}\cos\nu+\xi^2\textit{e}^{2\textit{i}\mu}))\textit{d}\xi$,
where $\mu$ and $\nu$ are real numbers. For any real number $\alpha$ and
harmonic function $f=h+\overline{g}$, define an analytic function
$f_{\alpha}:=h+\textit{e}^{-2\textit{i}\alpha}g$. Let $\mu_1$ and $\mu_2$
$(\mu_1+\mu_2=\mu)$ be real numbers, and $f=h+\overline{g}$ and
$F=H+\overline{G}$ be locally-univalent and sense-preserving harmonic functions
such that $f_{\mu_1}*F_{\mu_2}=\phi$. It is shown that the convolution $f*F$ is
univalent and convex in the direction of $-\mu$, provided it is locally
univalent and sense-preserving. Also, local-univalence of the above convolution
$f*F$ is shown for some specific analytic dilatations of $f$ and $F$.
Furthermore, if $g\equiv0$ and both the analytic functions $f_{\mu_1}$ and
$F_{\mu_2}$ are convex, then the convolution $f*F$ is shown to be convex. These
results extends the work done by Dorff \textit{et al.} to a larger class of
functions.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,263 | Optimal Non-uniform Deployments in Ultra-Dense Finite-Area Cellular Networks | Network densification and heterogenisation through the deployment of small
cellular access points (picocells and femtocells) are seen as key mechanisms in
handling the exponential increase in cellular data traffic. Modelling such
networks by leveraging tools from Stochastic Geometry has proven particularly
useful in understanding the fundamental limits imposed on network coverage and
capacity by co-channel interference. Most of these works however assume
infinite sized and uniformly distributed networks on the Euclidean plane. In
contrast, we study finite sized non-uniformly distributed networks, and find
the optimal non-uniform distribution of access points which maximises network
coverage for a given non-uniform distribution of mobile users, and vice versa.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,264 | Plan, Attend, Generate: Character-level Neural Machine Translation with Planning in the Decoder | We investigate the integration of a planning mechanism into an
encoder-decoder architecture with an explicit alignment for character-level
machine translation. We develop a model that plans ahead when it computes
alignments between the source and target sequences, constructing a matrix of
proposed future alignments and a commitment vector that governs whether to
follow or recompute the plan. This mechanism is inspired by the strategic
attentive reader and writer (STRAW) model. Our proposed model is end-to-end
trainable with fully differentiable operations. We show that it outperforms a
strong baseline on three character-level decoder neural machine translation on
WMT'15 corpus. Our analysis demonstrates that our model can compute
qualitatively intuitive alignments and achieves superior performance with fewer
parameters.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,265 | Deep Recurrent NMF for Speech Separation by Unfolding Iterative Thresholding | In this paper, we propose a novel recurrent neural network architecture for
speech separation. This architecture is constructed by unfolding the iterations
of a sequential iterative soft-thresholding algorithm (ISTA) that solves the
optimization problem for sparse nonnegative matrix factorization (NMF) of
spectrograms. We name this network architecture deep recurrent NMF (DR-NMF).
The proposed DR-NMF network has three distinct advantages. First, DR-NMF
provides better interpretability than other deep architectures, since the
weights correspond to NMF model parameters, even after training. This
interpretability also provides principled initializations that enable faster
training and convergence to better solutions compared to conventional random
initialization. Second, like many deep networks, DR-NMF is an order of
magnitude faster at test time than NMF, since computation of the network output
only requires evaluating a few layers at each time step. Third, when a limited
amount of training data is available, DR-NMF exhibits stronger generalization
and separation performance compared to sparse NMF and state-of-the-art
long-short term memory (LSTM) networks. When a large amount of training data is
available, DR-NMF achieves lower yet competitive separation performance
compared to LSTM networks.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,266 | Birth of isolated nested cylinders and limit cycles in 3D piecewise smooth vector fields with symmetry | Our start point is a 3D piecewise smooth vector field defined in two zones
and presenting a shared fold curve for the two smooth vector fields considered.
Moreover, these smooth vector fields are symmetric relative to the fold curve,
giving raise to a continuum of nested topological cylinders such that each
orthogonal section of these cylinders is filled by centers. First we prove that
the normal form considered represents a whole class of piecewise smooth vector
fields. After we perturb the initial model in order to obtain exactly
$\mathcal{L}$ invariant planes containing centers. A second perturbation of the
initial model also is considered in order to obtain exactly $k$ isolated
cylinders filled by periodic orbits. Finally, joining the two previous
bifurcations we are able to exhibit a model, preserving the symmetry relative
to the fold curve, and having exactly $k.\mathcal{L}$ limit cycles.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,267 | The difficulty of folding self-folding origami | Why is it difficult to refold a previously folded sheet of paper? We show
that even crease patterns with only one designed folding motion inevitably
contain an exponential number of `distractor' folding branches accessible from
a bifurcation at the flat state. Consequently, refolding a sheet requires
finding the ground state in a glassy energy landscape with an exponential
number of other attractors of higher energy, much like in models of protein
folding (Levinthal's paradox) and other NP-hard satisfiability (SAT) problems.
As in these problems, we find that refolding a sheet requires actuation at
multiple carefully chosen creases. We show that seeding successful folding in
this way can be understood in terms of sub-patterns that fold when cut out
(`folding islands'). Besides providing guidelines for the placement of active
hinges in origami applications, our results point to fundamental limits on the
programmability of energy landscapes in sheets.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,268 | Integrable Trotterization: Local Conservation Laws and Boundary Driving | We discuss a general procedure to construct an integrable real-time
trotterization of interacting lattice models. As an illustrative example we
consider a spin-$1/2$ chain, with continuous time dynamics described by the
isotropic ($XXX$) Heisenberg Hamiltonian. For periodic boundary conditions
local conservation laws are derived from an inhomogeneous transfer matrix and a
boost operator is constructed. In the continuous time limit these local charges
reduce to the known integrals of motion of the Heisenberg chain. In a simple
Kraus representation we also examine the nonequilibrium setting, where our
integrable cellular automaton is driven by stochastic processes at the
boundaries. We show explicitly, how an exact nonequilibrium steady state
density matrix can be written in terms of a staggered matrix product ansatz.
This simple trotterization scheme, in particular in the open system framework,
could prove to be a useful tool for experimental simulations of the lattice
models in terms of trapped ion and atom optics setups.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,269 | The Multivariate Hawkes Process in High Dimensions: Beyond Mutual Excitation | The Hawkes process is a class of point processes whose future depends on its
own history. Previous theoretical work on the Hawkes process is limited to the
case of a mutually-exciting process, in which a past event can only increase
the occurrence of future events. However, in neuronal networks and other
real-world applications, inhibitory relationships may be present. In this
paper, we develop a new approach for establishing the properties of the Hawkes
process without the restriction to mutual excitation. To this end, we employ a
thinning process representation and a coupling construction to bound the
dependence coefficient of the Hawkes process. Using recent developments on
weakly dependent sequences, we establish a concentration inequality for
second-order statistics of the Hawkes process. We apply this concentration
inequality in order to establish theoretical results for penalized regression
and clustering analysis in the high-dimensional regime. Our theoretical results
are corroborated by simulation studies and an application to a neuronal spike
train data set.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,270 | Aerial-Ground collaborative sensing: Third-Person view for teleoperation | Rapid deployment and operation are key requirements in time critical
application, such as Search and Rescue (SaR). Efficiently teleoperated ground
robots can support first-responders in such situations. However, first-person
view teleoperation is sub-optimal in difficult terrains, while a third-person
perspective can drastically increase teleoperation performance. Here, we
propose a Micro Aerial Vehicle (MAV)-based system that can autonomously provide
third-person perspective to ground robots. While our approach is based on local
visual servoing, it further leverages the global localization of several ground
robots to seamlessly transfer between these ground robots in GPS-denied
environments. Therewith one MAV can support multiple ground robots on a demand
basis. Furthermore, our system enables different visual detection regimes, and
enhanced operability, and return-home functionality. We evaluate our system in
real-world SaR scenarios.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,271 | A universal thin film model for Ginzburg-Landau energy with dipolar interaction | We present an analytical treatment of a three-dimensional variational model
of a system that exhibits a second-order phase transition in the presence of
dipolar interactions. Within the framework of Ginzburg-Landau theory, we
concentrate on the case in which the domain occupied by the sample has the
shape of a flat thin film and obtain a reduced two-dimensional, non-local
variational model that describes the energetics of the system in terms of the
order parameter averages across the film thickness. Namely, we show that the
reduced two-dimensional model is in a certain sense asymptotically equivalent
to the original three-dimensional model for small film thicknesses. Using this
asymptotic equivalence, we analyze two different thin film limits for the full
three-dimensional model via the methods of $\Gamma$-convergence applied to the
reduced two-dimensional model. In the first regime, in which the film thickness
vanishes while all other parameters remain fixed, we recover the local
two-dimensional Ginzburg-Landau model. On the other hand, when the film
thickness vanishes while the sample's lateral dimensions diverge at the right
rate, we show that the system exhibits a transition from homogeneous to
spatially modulated global energy minimizers. We identify a sharp threshold for
this transition.
| 0 | 1 | 1 | 0 | 0 | 0 |
20,272 | LOCATA challenge: speaker localization with a planar array | This document describes our submission to the 2018 LOCalization And TrAcking
(LOCATA) challenge (Tasks 1, 3, 5). We estimate the 3D position of a speaker
using the Global Coherence Field (GCF) computed from multiple microphone pairs
of a DICIT planar array. One of the main challenges when using such an array
with omnidirectional microphones is the front-back ambiguity, which is
particularly evident in Task 5. We address this challenge by post-processing
the peaks of the GCF and exploiting the attenuation introduced by the frame of
the array. Moreover, the intermittent nature of speech and the changing
orientation of the speaker make localization difficult. For Tasks 3 and 5, we
also employ a Particle Filter (PF) that favors the spatio-temporal continuity
of the localization results.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,273 | Less Is More: A Comprehensive Framework for the Number of Components of Ensemble Classifiers | The number of component classifiers chosen for an ensemble greatly impacts
the prediction ability. In this paper, we use a geometric framework for a
priori determining the ensemble size, which is applicable to most of existing
batch and online ensemble classifiers. There are only a limited number of
studies on the ensemble size examining Majority Voting (MV) and Weighted
Majority Voting (WMV). Almost all of them are designed for batch-mode, hardly
addressing online environments. Big data dimensions and resource limitations,
in terms of time and memory, make determination of ensemble size crucial,
especially for online environments. For the MV aggregation rule, our framework
proves that the more strong components we add to the ensemble, the more
accurate predictions we can achieve. For the WMV aggregation rule, our
framework proves the existence of an ideal number of components, which is equal
to the number of class labels, with the premise that components are completely
independent of each other and strong enough. While giving the exact definition
for a strong and independent classifier in the context of an ensemble is a
challenging task, our proposed geometric framework provides a theoretical
explanation of diversity and its impact on the accuracy of predictions. We
conduct a series of experimental evaluations to show the practical value of our
theorems and existing challenges.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,274 | Large-Scale Low-Rank Matrix Learning with Nonconvex Regularizers | Low-rank modeling has many important applications in computer vision and
machine learning. While the matrix rank is often approximated by the convex
nuclear norm, the use of nonconvex low-rank regularizers has demonstrated
better empirical performance. However, the resulting optimization problem is
much more challenging. Recent state-of-the-art requires an expensive full SVD
in each iteration. In this paper, we show that for many commonly-used nonconvex
low-rank regularizers, a cutoff can be derived to automatically threshold the
singular values obtained from the proximal operator. This allows such operator
being efficiently approximated by power method. Based on it, we develop a
proximal gradient algorithm (and its accelerated variant) with inexact proximal
splitting and prove that a convergence rate of O(1/T) where T is the number of
iterations is guaranteed. Furthermore, we show the proposed algorithm can be
well parallelized, which achieves nearly linear speedup w.r.t the number of
threads. Extensive experiments are performed on matrix completion and robust
principal component analysis, which shows a significant speedup over the
state-of-the-art. Moreover, the matrix solution obtained is more accurate and
has a lower rank than that of the nuclear norm regularizer.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,275 | Computational Eco-Systems for Handwritten Digits Recognition | Inspired by the importance of diversity in biological system, we built an
heterogeneous system that could achieve this goal. Our architecture could be
summarized in two basic steps. First, we generate a diverse set of
classification hypothesis using both Convolutional Neural Networks, currently
the state-of-the-art technique for this task, among with other traditional and
innovative machine learning techniques. Then, we optimally combine them through
Meta-Nets, a family of recently developed and performing ensemble methods.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,276 | Compressing networks with super nodes | Community detection is a commonly used technique for identifying groups in a
network based on similarities in connectivity patterns. To facilitate community
detection in large networks, we recast the network to be partitioned into a
smaller network of 'super nodes', each super node comprising one or more nodes
in the original network. To define the seeds of our super nodes, we apply the
'CoreHD' ranking from dismantling and decycling. We test our approach through
the analysis of two common methods for community detection: modularity
maximization with the Louvain algorithm and maximum likelihood optimization for
fitting a stochastic block model. Our results highlight that applying community
detection to the compressed network of super nodes is significantly faster
while successfully producing partitions that are more aligned with the local
network connectivity, more stable across multiple (stochastic) runs within and
between community detection algorithms, and overlap well with the results
obtained using the full network.
| 1 | 1 | 0 | 0 | 0 | 0 |
20,277 | Mode specific electronic friction in dissociative chemisorption on metal surfaces: H$_2$ on Ag(111) | Electronic friction and the ensuing nonadiabatic energy loss play an
important role in chemical reaction dynamics at metal surfaces. Using molecular
dynamics with electronic friction evaluated on-the-fly from Density Functional
Theory, we find strong mode dependence and a dominance of nonadiabatic energy
loss along the bond stretch coordinate for scattering and dissociative
chemisorption of H$_2$ on the Ag(111) surface. Exemplary trajectories with
varying initial conditions indicate that this mode-specificity translates into
modulated energy loss during a dissociative chemisorption event. Despite minor
nonadiabatic energy loss of about 5\%, the directionality of friction forces
induces dynamical steering that affects individual reaction outcomes,
specifically for low-incidence energies and vibrationally excited molecules.
Mode-specific friction induces enhanced loss of rovibrational rather than
translational energy and will be most visible in its effect on final energy
distributions in molecular scattering experiments.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,278 | The quantum auxiliary linear problem & quantum Darboux-Backlund transformations | We explore the notion of the quantum auxiliary linear problem and the
associated problem of quantum Backlund transformations (BT). In this context we
systematically construct the analogue of the classical formula that provides
the whole hierarchy of the time components of Lax pairs at the quantum level
for both closed and open integrable lattice models. The generic time evolution
operator formula is particularly interesting and novel at the quantum level
when dealing with systems with open boundary conditions. In the same frame we
show that the reflection K-matrix can also be viewed as a particular type of
BT, fixed at the boundaries of the system. The q-oscillator (q-boson) model, a
variant of the Ablowitz-Ladik model, is then employed as a paradigm to
illustrate the method. Particular emphasis is given to the time part of the
quantum BT as possible connections and applications to the problem of quantum
quenches as well as the time evolution of local quantum impurities are evident.
A discussion on the use of Bethe states as well as coherent states for the
study of the time evolution is also presented.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,279 | Simplified Minimal Gated Unit Variations for Recurrent Neural Networks | Recurrent neural networks with various types of hidden units have been used
to solve a diverse range of problems involving sequence data. Two of the most
recent proposals, gated recurrent units (GRU) and minimal gated units (MGU),
have shown comparable promising results on example public datasets. In this
paper, we introduce three model variants of the minimal gated unit (MGU) which
further simplify that design by reducing the number of parameters in the
forget-gate dynamic equation. These three model variants, referred to simply as
MGU1, MGU2, and MGU3, were tested on sequences generated from the MNIST dataset
and from the Reuters Newswire Topics (RNT) dataset. The new models have shown
similar accuracy to the MGU model while using fewer parameters and thus
lowering training expense. One model variant, namely MGU2, performed better
than MGU on the datasets considered, and thus may be used as an alternate to
MGU or GRU in recurrent neural networks.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,280 | Parallel G-duplex and C-duplex DNA with Uninterrupted Spines of AgI-Mediated Base Pairs | Hydrogen bonding between nucleobases produces diverse DNA structural motifs,
including canonical duplexes, guanine (G) quadruplexes and cytosine (C)
i-motifs. Incorporating metal-mediated base pairs into nucleic acid structures
can introduce new functionalities and enhanced stabilities. Here we
demonstrate, using mass spectrometry (MS), ion mobility spectrometry (IMS) and
fluorescence resonance energy transfer (FRET), that parallel-stranded
structures consisting of up to 20 G-Ag(I)-G contiguous base pairs are formed
when natural DNA sequences are mixed with silver cations in aqueous solution.
FRET indicates that duplexes formed by poly(cytosine) strands with 20
contiguous C-Ag(I)-C base pairs are also parallel. Silver-mediated G-duplexes
form preferentially over G-quadruplexes, and the ability of Ag+ to convert
G-quadruplexes into silver-paired duplexes may provide a new route to
manipulating these biologically relevant structures. IMS indicates that
G-duplexes are linear and more rigid than B-DNA. DFT calculations were used to
propose structures compatible with the IMS experiments. Such inexpensive,
defect-free and soluble DNA-based nanowires open new directions in the design
of novel metal-mediated DNA nanotechnology.
| 0 | 0 | 0 | 0 | 1 | 0 |
20,281 | Constraints on Vacuum Energy from Structure Formation and Nucleosynthesis | This paper derives an upper limit on the density $\rho_{\scriptstyle\Lambda}$
of dark energy based on the requirement that cosmological structure forms
before being frozen out by the eventual acceleration of the universe. By
allowing for variations in both the cosmological parameters and the strength of
gravity, the resulting constraint is a generalization of previous limits. The
specific parameters under consideration include the amplitude $Q$ of the
primordial density fluctuations, the Planck mass $M_{\rm pl}$, the
baryon-to-photon ratio $\eta$, and the density ratio $\Omega_M/\Omega_b$. In
addition to structure formation, we use considerations from stellar structure
and Big Bang Nucleosynthesis (BBN) to constrain these quantities. The resulting
upper limit on the dimensionless density of dark energy becomes
$\rho_{\scriptstyle\Lambda}/M_{\rm pl}^4<10^{-90}$, which is $\sim30$ orders of
magnitude larger than the value in our universe
$\rho_{\scriptstyle\Lambda}/M_{\rm pl}^4\sim10^{-120}$. This new limit is much
less restrictive than previous constraints because additional parameters are
allowed to vary. With these generalizations, a much wider range of universes
can develop cosmic structure and support observers. To constrain the
constituent parameters, new BBN calculations are carried out in the regime
where $\eta$ and $G=M_{\rm pl}^{-2}$ are much larger than in our universe. If
the BBN epoch were to process all of the protons into heavier elements, no
hydrogen would be left behind to make water, and the universe would not be
viable. However, our results show that some hydrogen is always left over, even
under conditions of extremely large $\eta$ and $G$, so that a wide range of
alternate universes are potentially habitable.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,282 | Dynamic Mobile Edge Caching with Location Differentiation | Mobile edge caching enables content delivery directly within the radio access
network, which effectively alleviates the backhaul burden and reduces
round-trip latency. To fully exploit the edge resources, the most popular
contents should be identified and cached. Observing that content popularity
varies greatly at different locations, to maximize local hit rate, this paper
proposes an online learning algorithm that dynamically predicts content hit
rate, and makes location-differentiated caching decisions. Specifically, a
linear model is used to estimate the future hit rate. Considering the
variations in user demand, a perturbation is added to the estimation to account
for uncertainty. The proposed learning algorithm requires no training phase,
and hence is adaptive to the time-varying content popularity profile.
Theoretical analysis indicates that the proposed algorithm asymptotically
approaches the optimal policy in the long term. Extensive simulations based on
real world traces show that, the proposed algorithm achieves higher hit rate
and better adaptiveness to content popularity fluctuation, compared with other
schemes.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,283 | Online Learning for Distribution-Free Prediction | We develop an online learning method for prediction, which is important in
problems with large and/or streaming data sets. We formulate the learning
approach using a covariance-fitting methodology, and show that the resulting
predictor has desirable computational and distribution-free properties: It is
implemented online with a runtime that scales linearly in the number of
samples; has a constant memory requirement; avoids local minima problems; and
prunes away redundant feature dimensions without relying on restrictive
assumptions on the data distribution. In conjunction with the split conformal
approach, it also produces distribution-free prediction confidence intervals in
a computationally efficient manner. The method is demonstrated on both real and
synthetic datasets.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,284 | Thermal transitions, pseudogap behavior and BCS-BEC crossover in Fermi-Fermi mixtures | We study the mass imbalanced Fermi-Fermi mixture within the framework of a
two-dimensional lattice fermion model. Based on the thermodynamic and species
dependent quasiparticle behavior we map out the finite temperature phase
diagram of this system and show that unlike the balanced Fermi superfluid there
are now two different pseudogap regimes as PG-I and PG-II. While within the
PG-I regime both the fermionic species are pseudogapped, PG-II corresponds to
the regime where pseudogap feature survives only in the light species. We
believe that the single particle spectral features that we discuss in this
paper are observable through the species resolved radio frequency spectroscopy
and momentum resolved photo emission spectroscopy measurements on systems such
as, 6$_{Li}$-40$_{K}$ mixture. We further investigate the interplay between the
population and mass imbalances and report that at a fixed population imbalance
the BCS-BEC crossover in a Fermi-Fermi mixture would require a critical
interaction (U$_{c}$), for the realization of the uniform superfluid state. The
effect of imbalance in mass on the exotic Fulde-Ferrell-Larkin-Ovchinnikov
(FFLO) superfluid phase has been probed in detail in terms of the thermodynamic
and quasiparticle behavior of this phase. It has been observed that in spite of
the s-wave symmetry of the pairing field a nodal superfluid gap is realized in
the LO regime. Our results on the various thermal scales and regimes are
expected to serve as benchmarks for the experimental observations on
6$_{Li}$-40$_{K}$ mixture.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,285 | Spectral Approximation for Ergodic CMV Operators with an Application to Quantum Walks | We establish concrete criteria for fully supported absolutely continuous
spectrum for ergodic CMV matrices and purely absolutely continuous spectrum for
limit-periodic CMV matrices. We proceed by proving several variational
estimates on the measure of the spectrum and the vanishing set of the Lyapunov
exponent for CMV matrices, which represent CMV analogues of results obtained
for Schrödinger operators due to Y.\ Last in the early 1990s. Having done so,
we combine those estimates with results from inverse spectral theory to obtain
purely absolutely continuous spectrum.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,286 | Quantitative Photoacoustic Imaging in the Acoustic Regime using SPIM | While in standard photoacoustic imaging the propagation of sound waves is
modeled by the standard wave equation, our approach is based on a generalized
wave equation with variable sound speed and material density, respectively. In
this paper we present an approach for photoacoustic imaging, which in addition
to recovering of the absorption density parameter, the imaging parameter of
standard photoacoustics, also allows to reconstruct the spatially varying sound
speed and density, respectively, of the medium. We provide analytical
reconstruction formulas for all three parameters based in a linearized model
based on single plane illumination microscopy (SPIM) techniques.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,287 | Integrated Deep and Shallow Networks for Salient Object Detection | Deep convolutional neural network (CNN) based salient object detection
methods have achieved state-of-the-art performance and outperform those
unsupervised methods with a wide margin. In this paper, we propose to integrate
deep and unsupervised saliency for salient object detection under a unified
framework. Specifically, our method takes results of unsupervised saliency
(Robust Background Detection, RBD) and normalized color images as inputs, and
directly learns an end-to-end mapping between inputs and the corresponding
saliency maps. The color images are fed into a Fully Convolutional Neural
Networks (FCNN) adapted from semantic segmentation to exploit high-level
semantic cues for salient object detection. Then the results from deep FCNN and
RBD are concatenated to feed into a shallow network to map the concatenated
feature maps to saliency maps. Finally, to obtain a spatially consistent
saliency map with sharp object boundaries, we fuse superpixel level saliency
map at multi-scale. Extensive experimental results on 8 benchmark datasets
demonstrate that the proposed method outperforms the state-of-the-art
approaches with a margin.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,288 | Feedback Capacity over Networks | In this paper, we investigate the fundamental limitations of feedback
mechanism in dealing with uncertainties for network systems. The study of
maximum capability of feedback control was pioneered in Xie and Guo (2000) for
scalar systems with nonparametric nonlinear uncertainty. In a network setting,
nodes with unknown and nonlinear dynamics are interconnected through a directed
interaction graph. Nodes can design feedback controls based on all available
information, where the objective is to stabilize the network state. Using
information structure and decision pattern as criteria, we specify three
categories of network feedback laws, namely the
global-knowledge/global-decision, network-flow/local-decision, and
local-flow/local-decision feedback. We establish a series of network capacity
characterizations for these three fundamental types of network control laws.
First of all, we prove that for global-knowledge/global-decision and
network-flow/local-decision control where nodes know the information flow
across the entire network, there exists a critical number
$\big(3/2+\sqrt{2}\big)/\|A_{\mathrm{G}}\|_\infty$, where $3/2+\sqrt{2}$ is as
known as the Xie-Guo constant and $A_{\mathrm{G}}$ is the network adjacency
matrix, defining exactly how much uncertainty in the node dynamics can be
overcome by feedback. Interestingly enough, the same feedback capacity can be
achieved under max-consensus enhanced local flows where nodes only observe
information flows from neighbors as well as extreme (max and min) states in the
network. Next, for local-flow/local-decision control, we prove that there
exists a structure-determined value being a lower bound of the network feedback
capacity. These results reveal the important connection between network
structure and fundamental capabilities of in-network feedback control.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,289 | Flexural phonons in supported graphene: from pinning to localization | We identify graphene layer on a disordered substrate as a possible system
where Anderson localization of phonons can be observed. Generally, observation
of localization for scattering waves is not simple, because the Rayleigh
scattering is inversely proportional to a high power of wavelength. The
situation is radically different for the out of plane vibrations, so-called
flexural phonons, scattered by pinning centers induced by a substrate. In this
case, the scattering time for vanishing wave vector tends to a finite limit.
One may, therefore, expect that physics of the flexural phonons exhibits
features characteristic for electron localization in two dimensions, albeit
without complications caused by the electron-electron interactions. We confirm
this idea by calculating statistical properties of the Anderson localization of
flexural phonons for a model of elastic sheet in the presence of the pinning
centers. Finally, we discuss possible manifestations of the flexural phonons,
including the localized ones, in the electronic thermal conductance.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,290 | Test them all, is it worth it? Assessing configuration sampling on the JHipster Web development stack | Many approaches for testing configurable software systems start from the same
assumption: it is impossible to test all configurations. This motivated the
definition of variability-aware abstractions and sampling techniques to cope
with large configuration spaces. Yet, there is no theoretical barrier that
prevents the exhaustive testing of all configurations by simply enumerating
them, if the effort required to do so remains acceptable. Not only this: we
believe there is lots to be learned by systematically and exhaustively testing
a configurable system. In this case study, we report on the first ever
endeavour to test all possible configurations of an industry-strength, open
source configurable software system, JHipster, a popular code generator for web
applications. We built a testing scaffold for the 26,000+ configurations of
JHipster using a cluster of 80 machines during 4 nights for a total of 4,376
hours (182 days) CPU time. We find that 35.70% configurations fail and we
identify the feature interactions that cause the errors. We show that sampling
strategies (like dissimilarity and 2-wise): (1) are more effective to find
faults than the 12 default configurations used in the JHipster continuous
integration; (2) can be too costly and exceed the available testing budget. We
cross this quantitative analysis with the qualitative assessment of JHipster's
lead developers.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,291 | Q-Learning Algorithm for VoLTE Closed-Loop Power Control in Indoor Small Cells | We propose a reinforcement learning (RL) based closed loop power control
algorithm for the downlink of the voice over LTE (VoLTE) radio bearer for an
indoor environment served by small cells. The main contributions of our paper
are to 1) use RL to solve performance tuning problems in an indoor cellular
network for voice bearers and 2) show that our derived lower bound loss in
effective signal to interference plus noise ratio due to neighboring cell
failure is sufficient for VoLTE power control purposes in practical cellular
networks. In our simulation, the proposed RL-based power control algorithm
significantly improves both voice retainability and mean opinion score compared
to current industry standards. The improvement is due to maintaining an
effective downlink signal to interference plus noise ratio against adverse
network operational issues and faults.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,292 | Ancient shrinking spherical interfaces in the Allen-Cahn flow | We consider the parabolic Allen-Cahn equation in $\mathbb{R}^n$, $n\ge 2$,
$$u_t= \Delta u + (1-u^2)u \quad \hbox{ in } \mathbb{R}^n \times (-\infty,
0].$$ We construct an ancient radially symmetric solution $u(x,t)$ with any
given number $k$ of transition layers between $-1$ and $+1$. At main order they
consist of $k$ time-traveling copies of $w$ with spherical interfaces distant
$O(\log |t| )$ one to each other as $t\to -\infty$. These interfaces are
resemble at main order copies of the {\em shrinking sphere} ancient solution to
mean the flow by mean curvature of surfaces: $|x| = \sqrt{- 2(n-1)t}$. More
precisely, if $w(s)$ denotes the heteroclinic 1-dimensional solution of $w'' +
(1-w^2)w=0$ $w(\pm \infty)= \pm 1$ given by $w(s) = \tanh \left(\frac
s{\sqrt{2}} \right) $ we have $$ u(x,t) \approx \sum_{j=1}^k
(-1)^{j-1}w(|x|-\rho_j(t)) - \frac 12 (1+ (-1)^{k}) \quad \hbox{ as } t\to
-\infty $$ where
$$\rho_j(t)=\sqrt{-2(n-1)t}+\frac{1}{\sqrt{2}}\left(j-\frac{k+1}{2}\right)\log\left(\frac
{|t|}{\log |t| }\right)+ O(1),\quad j=1,\ldots ,k.$$
| 0 | 0 | 1 | 0 | 0 | 0 |
20,293 | Measuring the Eccentricity of Items | The long-tail phenomenon tells us that there are many items in the tail.
However, not all tail items are the same. Each item acquires different kinds of
users. Some items are loved by the general public, while some items are
consumed by eccentric fans. In this paper, we propose a novel metric, item
eccentricity, to incorporate this difference between consumers of the items.
Eccentric items are defined as items that are consumed by eccentric users. We
used this metric to analyze two real-world datasets of music and movies and
observed the characteristics of items in terms of eccentricity. The results
showed that our defined eccentricity of an item does not change much over time,
and classified eccentric and noneccentric items present significantly distinct
characteristics. The proposed metric effectively separates the eccentric and
noneccentric items mixed in the tail, which could not be done with the previous
measures, which only consider the popularity of items.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,294 | Dataset: Rare Event Classification in Multivariate Time Series | A real-world dataset is provided from a pulp-and-paper manufacturing
industry. The dataset comes from a multivariate time series process. The data
contains a rare event of paper break that commonly occurs in the industry. The
data contains sensor readings at regular time-intervals (x's) and the event
label (y). The primary purpose of the data is thought to be building a
classification model for early prediction of the rare event. However, it can
also be used for multivariate time series data exploration and building other
supervised and unsupervised models.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,295 | Experimental verification of stopping-power prediction from single- and dual-energy computed tomography in biological tissues | An experimental setup for consecutive measurement of ion and x-ray absorption
in tissue or other materials is introduced. With this setup using a 3D-printed
sample container, the reference stopping-power ratio (SPR) of materials can be
measured with an uncertainty of below 0.1%. A total of 65 porcine and bovine
tissue samples were prepared for measurement, comprising five samples each of
13 tissue types representing about 80% of the total body mass (three different
muscle and fatty tissues, liver, kidney, brain, heart, blood, lung and bone).
Using a standard stoichiometric calibration for single-energy CT (SECT) as well
as a state-of-the-art dual-energy CT (DECT) approach, SPR was predicted for all
tissues and then compared to the measured reference. With the SECT approach,
the SPRs of all tissues were predicted with a mean error of (-0.84 $\pm$ 0.12)%
and a mean absolute error of (1.27 $\pm$ 0.12)%. In contrast, the DECT-based
SPR predictions were overall consistent with the measured reference with a mean
error of (-0.02 $\pm$ 0.15)% and a mean absolute error of (0.10 $\pm$ 0.15)%.
Thus, in this study, the potential of DECT to decrease range uncertainty could
be confirmed in biological tissue.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,296 | How to Beat Science and Influence People: Policy Makers and Propaganda in Epistemic Networks | In their recent book Merchants of Doubt [New York:Bloomsbury 2010], Naomi
Oreskes and Erik Conway describe the "tobacco strategy", which was used by the
tobacco industry to influence policy makers regarding the health risks of
tobacco products. The strategy involved two parts, consisting of (1) promoting
and sharing independent research supporting the industry's preferred position
and (2) funding additional research, but selectively publishing the results. We
introduce a model of the Tobacco Strategy, and use it to argue that both prongs
of the strategy can be extremely effective--even when policy makers rationally
update on all evidence available to them. As we elaborate, this model helps
illustrate the conditions under which the Tobacco Strategy is particularly
successful. In addition, we show how journalists engaged in "fair" reporting
can inadvertently mimic the effects of industry on public belief.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,297 | A complete and partial integrability technique of the Lorenz system | In this paper we deal with the well-known nonlinear Lorenz system that
describes the deterministic chaos phenomenon. We consider an interesting
problem with time-varying phenomena in quantum optics. Then we establish from
the motion equations the passage to the Lorenz system. Furthermore, we show
that the reduction to the third order non linear equation can be performed.
Therefore, the obtained differential equation can be analytically solved in
some special cases and transformed to Abel, Dufing, Painlevé and
generalized Emden-Fowler equations. So, a motivating technique that permitted a
complete and partial integrability of the Lorenz system is presented.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,298 | Comparision of the definitions of generalized solution of the Cauchy problem for quasi-linear equation | In preprint we consider and compare different definitions of generalized
solution of the Cauchy problem for 1d-scalar quasilinear equation (conservation
law). We start from the classical approaches goes back to I.M. Gelfand, O.A.
Oleinik, S.N. Kruzhkov and move to the modern finite-difference approximations
approaches belongs to A.A. Shananin and G.M. Henkin. We discuss the conditions
that provide definitions to be equivalent.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,299 | On The Complexity of Sparse Label Propagation | This paper investigates the computational complexity of sparse label
propagation which has been proposed recently for processing network structured
data. Sparse label propagation amounts to a convex optimization problem and
might be considered as an extension of basis pursuit from sparse vectors to
network structured datasets. Using a standard first-order oracle model, we
characterize the number of iterations for sparse label propagation to achieve a
prescribed accuracy. In particular, we derive an upper bound on the number of
iterations required to achieve a certain accuracy and show that this upper
bound is sharp for datasets having a chain structure (e.g., time series).
| 0 | 0 | 0 | 1 | 0 | 0 |
20,300 | Detecting Qualia in Natural and Artificial Agents | The Hard Problem of consciousness has been dismissed as an illusion. By
showing that computers are capable of experiencing, we show that they are at
least rudimentarily conscious with potential to eventually reach
superconsciousness. The main contribution of the paper is a test for confirming
certain subjective experiences in a tested agent. We follow with analysis of
benefits and problems with conscious machines and implications of such
capability on future of computing, machine rights and artificial intelligence
safety.
| 1 | 0 | 0 | 0 | 0 | 0 |
Subsets and Splits