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
|
|---|---|---|---|---|---|---|---|---|
20,801 | Orbifold equivalence: structure and new examples | Orbifold equivalence is a notion of symmetry that does not rely on group
actions. Among other applications, it leads to surprising connections between
hitherto unrelated singularities. While the concept can be defined in a very
general category-theoretic language, we focus on the most explicit setting in
terms of matrix factorisations, where orbifold equivalences arise from defects
with special properties. Examples are relatively difficult to construct, but we
uncover some structural features that distinguish orbifold equivalences -- most
notably a finite perturbation expansion. We use those properties to devise a
search algorithm, then present some new examples including Arnold
singularities.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,802 | Efficient Attention using a Fixed-Size Memory Representation | The standard content-based attention mechanism typically used in
sequence-to-sequence models is computationally expensive as it requires the
comparison of large encoder and decoder states at each time step. In this work,
we propose an alternative attention mechanism based on a fixed size memory
representation that is more efficient. Our technique predicts a compact set of
K attention contexts during encoding and lets the decoder compute an efficient
lookup that does not need to consult the memory. We show that our approach
performs on-par with the standard attention mechanism while yielding inference
speedups of 20% for real-world translation tasks and more for tasks with longer
sequences. By visualizing attention scores we demonstrate that our models learn
distinct, meaningful alignments.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,803 | Multiplicities of bifurcation sets of Pham singularities | The local multiplicities of the Maxwell sets in the spaces of versal
deformations of Pham holomorphic function singularities are calculated. A
similar calculation for some other bifurcation sets (generalized Stokes' sets)
defined by more complicated relations between the critical values is given.
Aplications to the complexity of algorithms enumerating topologically distinct
morsifications of complicated real function singularities are discussed.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,804 | Fast and scalable Gaussian process modeling with applications to astronomical time series | The growing field of large-scale time domain astronomy requires methods for
probabilistic data analysis that are computationally tractable, even with large
datasets. Gaussian Processes are a popular class of models used for this
purpose but, since the computational cost scales, in general, as the cube of
the number of data points, their application has been limited to small
datasets. In this paper, we present a novel method for Gaussian Process
modeling in one-dimension where the computational requirements scale linearly
with the size of the dataset. We demonstrate the method by applying it to
simulated and real astronomical time series datasets. These demonstrations are
examples of probabilistic inference of stellar rotation periods, asteroseismic
oscillation spectra, and transiting planet parameters. The method exploits
structure in the problem when the covariance function is expressed as a mixture
of complex exponentials, without requiring evenly spaced observations or
uniform noise. This form of covariance arises naturally when the process is a
mixture of stochastically-driven damped harmonic oscillators -- providing a
physical motivation for and interpretation of this choice -- but we also
demonstrate that it can be a useful effective model in some other cases. We
present a mathematical description of the method and compare it to existing
scalable Gaussian Process methods. The method is fast and interpretable, with a
range of potential applications within astronomical data analysis and beyond.
We provide well-tested and documented open-source implementations of this
method in C++, Python, and Julia.
| 0 | 1 | 0 | 1 | 0 | 0 |
20,805 | Toward perfect reads: self-correction of short reads via mapping on de Bruijn graphs | Motivations Short-read accuracy is important for downstream analyses such as
genome assembly and hybrid long-read correction. Despite much work on
short-read correction, present-day correctors either do not scale well on large
data sets or consider reads as mere suites of k-mers, without taking into
account their full-length read information. Results We propose a new method to
correct short reads using de Bruijn graphs, and implement it as a tool called
Bcool. As a first st ep, Bcool constructs a compacted de Bruijn graph from the
reads. This graph is filtered on the basis of k-mer abundance then of unitig
abundance, thereby removing from most sequencing errors. The cleaned graph is
then used as a reference on which the reads are mapped to correct them. We show
that this approach yields more accurate reads than k-mer-spectrum correctors
while being scalable to human-size genomic datasets and beyond. Availability
and Implementation The implementation is open source and available at http:
//github.com/Malfoy/BCOOL under the Affero GPL license. Contact Antoine
Limasset [email protected] & Jean-François Flot [email protected] &
Pierre Peterlongo [email protected]
| 1 | 0 | 0 | 0 | 0 | 0 |
20,806 | Learning Pain from Action Unit Combinations: A Weakly Supervised Approach via Multiple Instance Learning | Patient pain can be detected highly reliably from facial expressions using a
set of facial muscle-based action units (AUs) defined by the Facial Action
Coding System (FACS). A key characteristic of facial expression of pain is the
simultaneous occurrence of pain-related AU combinations, whose automated
detection would be highly beneficial for efficient and practical pain
monitoring. Existing general Automated Facial Expression Recognition (AFER)
systems prove inadequate when applied specifically for detecting pain as they
either focus on detecting individual pain-related AUs but not on combinations
or they seek to bypass AU detection by training a binary pain classifier
directly on pain intensity data but are limited by lack of enough labeled data
for satisfactory training. In this paper, we propose a new approach that mimics
the strategy of human coders of decoupling pain detection into two consecutive
tasks: one performed at the individual video-frame level and the other at
video-sequence level. Using state-of-the-art AFER tools to detect single AUs at
the frame level, we propose two novel data structures to encode AU combinations
from single AU scores. Two weakly supervised learning frameworks namely
multiple instance learning (MIL) and multiple clustered instance learning
(MCIL) are employed corresponding to each data structure to learn pain from
video sequences. Experimental results show an 87% pain recognition accuracy
with 0.94 AUC (Area Under Curve) on the UNBC-McMaster Shoulder Pain Expression
dataset. Tests on long videos in a lung cancer patient video dataset
demonstrates the potential value of the proposed system for pain monitoring in
clinical settings.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,807 | A Causal And-Or Graph Model for Visibility Fluent Reasoning in Tracking Interacting Objects | Tracking humans that are interacting with the other subjects or environment
remains unsolved in visual tracking, because the visibility of the human of
interests in videos is unknown and might vary over time. In particular, it is
still difficult for state-of-the-art human trackers to recover complete human
trajectories in crowded scenes with frequent human interactions. In this work,
we consider the visibility status of a subject as a fluent variable, whose
change is mostly attributed to the subject's interaction with the surrounding,
e.g., crossing behind another object, entering a building, or getting into a
vehicle, etc. We introduce a Causal And-Or Graph (C-AOG) to represent the
causal-effect relations between an object's visibility fluent and its
activities, and develop a probabilistic graph model to jointly reason the
visibility fluent change (e.g., from visible to invisible) and track humans in
videos. We formulate this joint task as an iterative search of a feasible
causal graph structure that enables fast search algorithm, e.g., dynamic
programming method. We apply the proposed method on challenging video sequences
to evaluate its capabilities of estimating visibility fluent changes of
subjects and tracking subjects of interests over time. Results with comparisons
demonstrate that our method outperforms the alternative trackers and can
recover complete trajectories of humans in complicated scenarios with frequent
human interactions.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,808 | Exact Affine Counter Automata | We introduce an affine generalization of counter automata, and analyze their
ability as well as affine finite automata. Our contributions are as follows. We
show that there is a language that can be recognized by exact realtime affine
counter automata but by neither 1-way deterministic pushdown automata nor
realtime deterministic k-counter automata. We also show that a certain promise
problem, which is conjectured not to be solved by two-way quantum finite
automata in polynomial time, can be solved by Las Vegas affine finite automata.
Lastly, we show that how a counter helps for affine finite automata by showing
that the language MANYTWINS, which is conjectured not to be recognized by
affine, quantum or classical finite state models in polynomial time, can be
recognized by affine counter automata with one-sided bounded-error in realtime.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,809 | Preorder characterizations of lower separation axioms and their applications to foliations and flows | In this paper, we characterize several lower separation axioms $C_0, C_D$,
$C_R$, $C_N$, $\lambda$-space, nested, $S_{YS}$, $S_{YY}$, $S_{YS}$, and
$S_{\delta}$ using pre-order. To analyze topological properties of (resp.
dynamical systems) foliations, we introduce notions of topology (resp.
dynamical systems) for foliations. Then proper (resp. compact, minimal,
recurrent) foliations are characterized by separation axioms. Conversely, lower
separation axioms are interpreted into the condition for foliations and several
relations of them are described. Moreover, we introduce some notions for
topologies from dynamical systems and foliation theory.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,810 | Convergence of row sequences of simultaneous Padé-Faber approximants | We consider row sequences of vector valued Padé-Faber approximants
(simultaneous Padé-Faber approximants) and prove a Montessus de Ballore
type theorem.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,811 | A Deep Convolutional Neural Network for Background Subtraction | In this work, we present a novel background subtraction system that uses a
deep Convolutional Neural Network (CNN) to perform the segmentation. With this
approach, feature engineering and parameter tuning become unnecessary since the
network parameters can be learned from data by training a single CNN that can
handle various video scenes. Additionally, we propose a new approach to
estimate background model from video. For the training of the CNN, we employed
randomly 5 percent video frames and their ground truth segmentations taken from
the Change Detection challenge 2014(CDnet 2014). We also utilized
spatial-median filtering as the post-processing of the network outputs. Our
method is evaluated with different data-sets, and the network outperforms the
existing algorithms with respect to the average ranking over different
evaluation metrics. Furthermore, due to the network architecture, our CNN is
capable of real time processing.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,812 | A second main theorem for holomorphic curve intersecting hypersurfaces | In this paper, we establish a second main theorem for holomorphic curve
intersecting hypersurfaces in general position in projective space with level
of truncation. As an application, we reduce the number hypersurfaces in
uniqueness problem for holomorphic curve of authors before.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,813 | Partially hyperbolic diffeomorphisms with one-dimensional neutral center on 3-manifolds | We prove that for any partially hyperbolic diffeomorphism with one
dimensional neutral center on a 3-manifold, the center stable and center
unstable foliations are complete; moreover, each leaf of center stable and
center unstable foliations is a cylinder, a M$\ddot{o}$bius band or a plane.
Further properties of the Bonatti-Parwani-Potrie type of partially hyperbolic
diffeomorphisms are studied. Such examples are obtained by composing the time
$m$-map (for $m>0$ large) of a non-transitive Anosov flow $\phi_t$ on an
orientable 3-manifold with Dehn twists along some transverse tori, and the
examples are partially hyperbolic with one-dimensional neutral center. We prove
that the center foliation gives a topologically Anosov flow which is
topologically equivalent to $\phi_t$. We also prove that for the precise
example constructed by Bonatti-Parwani-Potrie, the center stable and center
unstable foliations are robustly complete.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,814 | Audio-Visual Speech Enhancement based on Multimodal Deep Convolutional Neural Network | Speech enhancement (SE) aims to reduce noise in speech signals. Most SE
techniques focus on addressing audio information only. In this work, inspired
by multimodal learning, which utilizes data from different modalities, and the
recent success of convolutional neural networks (CNNs) in SE, we propose an
audio-visual deep CNN (AVDCNN) SE model, which incorporates audio and visual
streams into a unified network model. In the proposed AVDCNN SE model, audio
and visual data are first processed using individual CNNs, and then, fused into
a joint network to generate enhanced speech at the output layer. The AVDCNN
model is trained in an end-to-end manner, and parameters are jointly learned
through back-propagation. We evaluate enhanced speech using five objective
criteria. Results show that the AVDCNN yields notably better performance,
compared with an audio-only CNN-based SE model and two conventional SE
approaches, confirming the effectiveness of integrating visual information into
the SE process.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,815 | Finding structure in the dark: coupled dark energy, weak lensing, and the mildly nonlinear regime | We reexamine interactions between the dark sectors of cosmology, with a focus
on robust constraints that can be obtained using only mildly nonlinear scales.
While it is well known that couplings between dark matter and dark energy can
be constrained to the percent level when including the full range of scales
probed by future optical surveys, calibrating matter power spectrum emulators
to all possible choices of potentials and couplings requires many
computationally expensive n-body simulations. Here we show that lensing and
clustering of galaxies in combination with the Cosmic Microwave Background
(CMB) is capable of probing the dark sector coupling to the few percent level
for a given class of models, using only linear and quasi-linear Fourier modes.
These scales can, in principle, be described by semi-analytical techniques such
as the effective field theory of large-scale structure.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,816 | Semantic Autoencoder for Zero-Shot Learning | Existing zero-shot learning (ZSL) models typically learn a projection
function from a feature space to a semantic embedding space (e.g.~attribute
space). However, such a projection function is only concerned with predicting
the training seen class semantic representation (e.g.~attribute prediction) or
classification. When applied to test data, which in the context of ZSL contains
different (unseen) classes without training data, a ZSL model typically suffers
from the project domain shift problem. In this work, we present a novel
solution to ZSL based on learning a Semantic AutoEncoder (SAE). Taking the
encoder-decoder paradigm, an encoder aims to project a visual feature vector
into the semantic space as in the existing ZSL models. However, the decoder
exerts an additional constraint, that is, the projection/code must be able to
reconstruct the original visual feature. We show that with this additional
reconstruction constraint, the learned projection function from the seen
classes is able to generalise better to the new unseen classes. Importantly,
the encoder and decoder are linear and symmetric which enable us to develop an
extremely efficient learning algorithm. Extensive experiments on six benchmark
datasets demonstrate that the proposed SAE outperforms significantly the
existing ZSL models with the additional benefit of lower computational cost.
Furthermore, when the SAE is applied to supervised clustering problem, it also
beats the state-of-the-art.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,817 | An Information-Theoretic Optimality Principle for Deep Reinforcement Learning | We methodologically address the problem of Q-value overestimation in deep
reinforcement learning to handle high-dimensional state spaces efficiently. By
adapting concepts from information theory, we introduce an intrinsic penalty
signal encouraging reduced Q-value estimates. The resultant algorithm
encompasses a wide range of learning outcomes containing deep Q-networks as a
special case. Different learning outcomes can be demonstrated by tuning a
Lagrange multiplier accordingly. We furthermore propose a novel scheduling
scheme for this Lagrange multiplier to ensure efficient and robust learning. In
experiments on Atari, our algorithm outperforms other algorithms (e.g. deep and
double deep Q-networks) in terms of both game-play performance and sample
complexity. These results remain valid under the recently proposed dueling
architecture.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,818 | Noise-synchronizability of opinion dynamics | With the analysis of noise-induced synchronization of opinion dynamics with
bounded confidence (BC), a natural and fundamental question is what opinion
structures can be synchronized by noise. In the traditional Hegselmann-Krause
(HK) model, each agent examines the opinion values of all the other ones and
then choose neighbors to update its own opinion according to the BC scheme. In
reality, people are more likely to interchange opinions with only some
individuals, resulting in a predetermined local discourse relationship as
introduced by the DeGroot model. In this paper, we consider an opinion dynamics
that combines the schemes of BC and local discourse topology and investigate
its synchronization induced by noise. The new model endows the heterogeneous HK
model with a time-varying discourse topology. With the proposed definition of
noise-synchronizability, it is shown that the compound noisy model is almost
surely noise-synchronizable if and only if the time-varying discourse graph is
uniformly jointly connected, taking the noise-induced synchronization of the
classical heterogeneous HK model as a special case. As a natural implication,
the result for the first time builds the equivalence between the connectivity
of discourse graph and the beneficial effect of noise for opinion consensus.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,819 | Conformer-selection by matter-wave interference | We establish that matter-wave interference at near-resonant ultraviolet
optical gratings can be used to spatially separate individual conformers of
complex molecules. Our calculations show that the conformational purity of the
prepared beam can be close to 100% and that all molecules remain in their
electronic ground state. The proposed technique is independent of the dipole
moment and the spin of the molecule and thus paves the way for
structure-sensitive experiments with hydrocarbons and biomolecules, such as
neurotransmitters and hormones, which evaded conformer-pure isolation so far
| 0 | 1 | 0 | 0 | 0 | 0 |
20,820 | Building a Neural Machine Translation System Using Only Synthetic Parallel Data | Recent works have shown that synthetic parallel data automatically generated
by translation models can be effective for various neural machine translation
(NMT) issues. In this study, we build NMT systems using only synthetic parallel
data. As an efficient alternative to real parallel data, we also present a new
type of synthetic parallel corpus. The proposed pseudo parallel data are
distinct from previous works in that ground truth and synthetic examples are
mixed on both sides of sentence pairs. Experiments on Czech-German and
French-German translations demonstrate the efficacy of the proposed pseudo
parallel corpus, which shows not only enhanced results for bidirectional
translation tasks but also substantial improvement with the aid of a ground
truth real parallel corpus.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,821 | Learning to Predict Indoor Illumination from a Single Image | We propose an automatic method to infer high dynamic range illumination from
a single, limited field-of-view, low dynamic range photograph of an indoor
scene. In contrast to previous work that relies on specialized image capture,
user input, and/or simple scene models, we train an end-to-end deep neural
network that directly regresses a limited field-of-view photo to HDR
illumination, without strong assumptions on scene geometry, material
properties, or lighting. We show that this can be accomplished in a three step
process: 1) we train a robust lighting classifier to automatically annotate the
location of light sources in a large dataset of LDR environment maps, 2) we use
these annotations to train a deep neural network that predicts the location of
lights in a scene from a single limited field-of-view photo, and 3) we
fine-tune this network using a small dataset of HDR environment maps to predict
light intensities. This allows us to automatically recover high-quality HDR
illumination estimates that significantly outperform previous state-of-the-art
methods. Consequently, using our illumination estimates for applications like
3D object insertion, we can achieve results that are photo-realistic, which is
validated via a perceptual user study.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,822 | $H^\infty$-calculus for semigroup generators on BMO | We prove that the negative infinitesimal generator $L$ of a semigroup of
positive contractions on $L^\infty$ has a bounded $H^\infty(S_\eta^0)$-calculus
on BMO$(\sqrt L)$ for any angle $\eta>\pi/2$, provided the semigroup satisfies
Bakry-Emry's $\Gamma_2 $ criterion. Our arguments only rely on the properties
of the underlying semigroup and works well in the noncommutative setting. A key
ingredient of our argument is a quasi monotone property for the subordinated
semigroup $T_{t,\alpha}=e^{-tL^\alpha},0<\alpha<1$, that is proved in the first
half of the article.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,823 | A Bayesian Game without epsilon equilibria | We present a three player Bayesian game for which there is no epsilon
equilibria in Borel measurable strategies for small enough epsilon, however
there are non-measurable equilibria.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,824 | Towards Robust Interpretability with Self-Explaining Neural Networks | Most recent work on interpretability of complex machine learning models has
focused on estimating $\textit{a posteriori}$ explanations for previously
trained models around specific predictions. $\textit{Self-explaining}$ models
where interpretability plays a key role already during learning have received
much less attention. We propose three desiderata for explanations in general --
explicitness, faithfulness, and stability -- and show that existing methods do
not satisfy them. In response, we design self-explaining models in stages,
progressively generalizing linear classifiers to complex yet architecturally
explicit models. Faithfulness and stability are enforced via regularization
specifically tailored to such models. Experimental results across various
benchmark datasets show that our framework offers a promising direction for
reconciling model complexity and interpretability.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,825 | Deep Learning Scaling is Predictable, Empirically | Deep learning (DL) creates impactful advances following a virtuous recipe:
model architecture search, creating large training data sets, and scaling
computation. It is widely believed that growing training sets and models should
improve accuracy and result in better products. As DL application domains grow,
we would like a deeper understanding of the relationships between training set
size, computational scale, and model accuracy improvements to advance the
state-of-the-art.
This paper presents a large scale empirical characterization of
generalization error and model size growth as training sets grow. We introduce
a methodology for this measurement and test four machine learning domains:
machine translation, language modeling, image processing, and speech
recognition. Our empirical results show power-law generalization error scaling
across a breadth of factors, resulting in power-law exponents---the "steepness"
of the learning curve---yet to be explained by theoretical work. Further, model
improvements only shift the error but do not appear to affect the power-law
exponent. We also show that model size scales sublinearly with data size. These
scaling relationships have significant implications on deep learning research,
practice, and systems. They can assist model debugging, setting accuracy
targets, and decisions about data set growth. They can also guide computing
system design and underscore the importance of continued computational scaling.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,826 | Coupled Self-Organized Hydrodynamics and Stokes models for suspensions of active particles | We derive macroscopic dynamics for self-propelled particles in a fluid. The
starting point is a coupled Vicsek-Stokes system. The Vicsek model describes
self-propelled agents interacting through alignment. It provides a
phenomenological description of hydrodynamic interactions between agents at
high density. Stokes equations describe a low Reynolds number fluid. These two
dynamics are coupled by the interaction between the agents and the fluid. The
fluid contributes to rotating the particles through Jeffery's equation.
Particle self-propulsion induces a force dipole on the fluid. After
coarse-graining we obtain a coupled Self-Organised Hydrodynamics (SOH)-Stokes
system. We perform a linear stability analysis for this system which shows that
both pullers and pushers have unstable modes. We conclude by providing
extensions of the Vicsek-Stokes model including short-distance repulsion,
finite particle inertia and finite Reynolds number fluid regime.
| 0 | 1 | 1 | 0 | 0 | 0 |
20,827 | Tunable Spin-Orbit Torques in Cu-Ta Binary Alloy Heterostructures | The spin Hall effect (SHE) is found to be strong in heavy transition metals
(HM), such as Ta and W, in their amorphous and/or high resistivity form. In
this work, we show that by employing a Cu-Ta binary alloy as buffer layer in an
amorphous Cu$_{100-x}$Ta$_{x}$-based magnetic heterostructure with
perpendicular magnetic anisotropy (PMA), the SHE-induced damping-like
spin-orbit torque (DL-SOT) efficiency $|\xi_{DL}|$ can be linearly tuned by
adjusting the buffer layer resistivity. Current-induced SOT switching can also
be achieved in these Cu$_{100-x}$Ta$_{x}$-based magnetic heterostructures, and
we find the switching behavior better explained by a SOT-assisted domain wall
propagation picture. Through systematic studies on Cu$_{100-x}$Ta$_{x}$-based
samples with various compositions, we determine the lower bound of spin Hall
conductivity
$|\sigma_{SH}|\approx2.02\times10^{4}[\hbar/2e]\Omega^{-1}\cdot\operatorname{m}^{-1}$
in the Ta-rich regime. Based on the idea of resistivity tuning, we further
demonstrate that $|\xi_{DL}|$ can be enhanced from 0.087 for pure Ta to 0.152
by employing a resistive TaN buffer layer.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,828 | On Properties of Nests: Some Answers and Questions | By considering nests on a given space, we explore order-theoretical and
topological properties that are closely related to the structure of a nest. In
particular, we see how subbases given by two dual nests can be an indicator of
how close or far are the properties of the space from the structure of a
linearly ordered space. Having in mind that the term interlocking nest is a key
tool to a general solution of the orderability problem, we give a
characterization of interlocking nest via closed sets in the Alexandroff
topology and via lower sets, respectively. We also characterize bounded subsets
of a given set in terms of nests and, finally, we explore the possibility of
characterizing topological groups via properties of nests. All sections are
followed by a number of open questions, which may give new directions to the
orderability problem.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,829 | Differential Characters of Drinfeld Modules and de Rham Cohomology | We introduce differential characters of Drinfeld modules. These are
function-field analogues of Buium's p-adic differential characters of elliptic
curves and of Manin's differential characters of elliptic curves in
differential algebra, both of which have had notable Diophantine applications.
We determine the structure of the group of differential characters. This shows
the existence of a family of interesting differential modular functions on the
moduli of Drinfeld modules. It also leads to a canonical $F$-crystal equipped
with a map to the de Rham cohomology of the Drinfeld module. This $F$-crystal
is of a differential-algebraic nature, and the relation to the classical
cohomological realizations is presently not clear.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,830 | Strong submeasures and several applications | A strong submeasure on a compact metric space X is a sub-linear and bounded
operator on the space of continuous functions on X. A strong submeasure is
positive if it is non-decreasing. By Hahn-Banach theorem, a positive strong
submeasure is the supremum of a non-empty collection of measures whose masses
are uniformly bounded from above.
We give several applications of strong submeasures in various diverse topics,
thus illustrate the usefulness of this classical but largely overlooked notion.
The applications include:
- Pullback and pushforward of all measures by meromorphic selfmaps of compact
complex varieties.
- The existence of invariant positive strong submeasures for meromorphic maps
between compact complex varieties, a notion of entropy for such submeasures
(which coincide with the classical ones in good cases) and a version of the
Variation Principle.
- Intersection of every positive closed (1,1) currents on compact Kähler
manifolds. Explicit calculations are given for self-intersection of the current
of integration of some curves $C$ in a compact Kähler surface where the
self-intersection in cohomology is negative.
All of these points are new and have not been previously given in work by
other authors. In addition, we will apply the same ideas to entropy of
transcendental maps of $\mathbb{C}$ and $\mathbb{C}^2$.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,831 | Understanding Career Progression in Baseball Through Machine Learning | Professional baseball players are increasingly guaranteed expensive long-term
contracts, with over 70 deals signed in excess of \$90 million, mostly in the
last decade. These are substantial sums compared to a typical franchise
valuation of \$1-2 billion. Hence, the players to whom a team chooses to give
such a contract can have an enormous impact on both competitiveness and profit.
Despite this, most published approaches examining career progression in
baseball are fairly simplistic. We applied four machine learning algorithms to
the problem and soundly improved upon existing approaches, particularly for
batting data.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,832 | Testing the validity of the local and global GKLS master equations on an exactly solvable model | When deriving a master equation for a multipartite weakly-interacting open
quantum systems, dissipation is often addressed \textit{locally} on each
component, i.e. ignoring the coherent couplings, which are later added `by
hand'. Although simple, the resulting local master equation (LME) is known to
be thermodynamically inconsistent. Otherwise, one may always obtain a
consistent \textit{global} master equation (GME) by working on the energy basis
of the full interacting Hamiltonian. Here, we consider a two-node `quantum
wire' connected to two heat baths. The stationary solution of the LME and GME
are obtained and benchmarked against the exact result. Importantly, in our
model, the validity of the GME is constrained by the underlying secular
approximation. Whenever this breaks down (for resonant weakly-coupled nodes),
we observe that the LME, in spite of being thermodynamically flawed: (a)
predicts the correct steady state, (b) yields the exact asymptotic heat
currents, and (c) reliably reflects the correlations between the nodes. In
contrast, the GME fails at all three tasks. Nonetheless, as the inter-node
coupling grows, the LME breaks down whilst the GME becomes correct. Hence, the
global and local approach may be viewed as \textit{complementary} tools, best
suited to different parameter regimes.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,833 | Unsupervised Contact Learning for Humanoid Estimation and Control | This work presents a method for contact state estimation using fuzzy
clustering to learn contact probability for full, six-dimensional humanoid
contacts. The data required for training is solely from proprioceptive sensors
- endeffector contact wrench sensors and inertial measurement units (IMUs) -
and the method is completely unsupervised. The resulting cluster means are used
to efficiently compute the probability of contact in each of the six
endeffector degrees of freedom (DoFs) independently. This clustering-based
contact probability estimator is validated in a kinematics-based base state
estimator in a simulation environment with realistic added sensor noise for
locomotion over rough, low-friction terrain on which the robot is subject to
foot slip and rotation. The proposed base state estimator which utilizes these
six DoF contact probability estimates is shown to perform considerably better
than that which determines kinematic contact constraints purely based on
measured normal force.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,834 | Learning to Sequence Robot Behaviors for Visual Navigation | Recent literature in the robotics community has focused on learning robot
behaviors that abstract out lower-level details of robot control. To fully
leverage the efficacy of such behaviors, it is necessary to select and sequence
them to achieve a given task. In this paper, we present an approach to both
learn and sequence robot behaviors, applied to the problem of visual navigation
of mobile robots. We construct a layered representation of control policies
composed of low- level behaviors and a meta-level policy. The low-level
behaviors enable the robot to locomote in a particular environment while
avoiding obstacles, and the meta-level policy actively selects the low-level
behavior most appropriate for the current situation based purely on visual
feedback. We demonstrate the effectiveness of our method on three simulated
robot navigation tasks: a legged hexapod robot which must successfully traverse
varying terrain, a wheeled robot which must navigate a maze-like course while
avoiding obstacles, and finally a wheeled robot navigating in the presence of
dynamic obstacles. We show that by learning control policies in a layered
manner, we gain the ability to successfully traverse new compound environments
composed of distinct sub-environments, and outperform both the low-level
behaviors in their respective sub-environments, as well as a hand-crafted
selection of low-level policies on these compound environments.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,835 | Evaluation complexity bounds for smooth constrained nonlinear optimisation using scaled KKT conditions, high-order models and the criticality measure $χ$ | Evaluation complexity for convexly constrained optimization is considered and
it is shown first that the complexity bound of $O(\epsilon^{-3/2})$ proved by
Cartis, Gould and Toint (IMAJNA 32(4) 2012, pp.1662-1695) for computing an
$\epsilon$-approximate first-order critical point can be obtained under
significantly weaker assumptions. Moreover, the result is generalized to the
case where high-order derivatives are used, resulting in a bound of
$O(\epsilon^{-(p+1)/p})$ evaluations whenever derivatives of order $p$ are
available. It is also shown that the bound of
$O(\epsilon_P^{-1/2}\epsilon_D^{-3/2})$ evaluations ($\epsilon_P$ and
$\epsilon_D$ being primal and dual accuracy thresholds) suggested by Cartis,
Gould and Toint (SINUM, 2015) for the general nonconvex case involving both
equality and inequality constraints can be generalized to a bound of
$O(\epsilon_P^{-1/p}\epsilon_D^{-(p+1)/p})$ evaluations under similarly
weakened assumptions. This paper is variant of a companion report (NTR-11-2015,
University of Namur, Belgium) which uses a different first-order criticality
measure to obtain the same complexity bounds.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,836 | One level density of low-lying zeros of quadratic and quartic Hecke $L$-functions | In this paper, we prove some one level density results for the low-lying
zeros of famliies of quadratic and quartic Hecke $L$-functions of the Gaussian
field. As corollaries, we deduce that, respectively, at least $94.27 \%$ and
$5\%$ of the members of the quadratic family and the quartic family do not
vanish at the central point.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,837 | An Online Convex Optimization Approach to Dynamic Network Resource Allocation | Existing approaches to online convex optimization (OCO) make sequential
one-slot-ahead decisions, which lead to (possibly adversarial) losses that
drive subsequent decision iterates. Their performance is evaluated by the
so-called regret that measures the difference of losses between the online
solution and the best yet fixed overall solution in hindsight. The present
paper deals with online convex optimization involving adversarial loss
functions and adversarial constraints, where the constraints are revealed after
making decisions, and can be tolerable to instantaneous violations but must be
satisfied in the long term. Performance of an online algorithm in this setting
is assessed by: i) the difference of its losses relative to the best dynamic
solution with one-slot-ahead information of the loss function and the
constraint (that is here termed dynamic regret); and, ii) the accumulated
amount of constraint violations (that is here termed dynamic fit). In this
context, a modified online saddle-point (MOSP) scheme is developed, and proved
to simultaneously yield sub-linear dynamic regret and fit, provided that the
accumulated variations of per-slot minimizers and constraints are sub-linearly
growing with time. MOSP is also applied to the dynamic network resource
allocation task, and it is compared with the well-known stochastic dual
gradient method. Under various scenarios, numerical experiments demonstrate the
performance gain of MOSP relative to the state-of-the-art.
| 1 | 0 | 1 | 1 | 0 | 0 |
20,838 | Learning causal Bayes networks using interventional path queries in polynomial time and sample complexity | Causal discovery from empirical data is a fundamental problem in many
scientific domains. Observational data allows for identifiability only up to
Markov equivalence class. In this paper we first propose a polynomial time
algorithm for learning the exact correctly-oriented structure of the transitive
reduction of any causal Bayesian networks with high probability, by using
interventional path queries. Each path query takes as input an origin node and
a target node, and answers whether there is a directed path from the origin to
the target. This is done by intervening the origin node and observing samples
from the target node. We theoretically show the logarithmic sample complexity
for the size of interventional data per path query, for continuous and discrete
networks. We further extend our work to learn the transitive edges using
logarithmic sample complexity (albeit in time exponential in the maximum number
of parents for discrete networks). This allows us to learn the full network. We
also provide an analysis of imperfect interventions.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,839 | The Fredholm alternative for the $p$-Laplacian in exterior domains | We investigate the Fredholm alternative for the $p$-Laplacian in an exterior
domain which is the complement of the closed unit ball in $\mathbb{R}^N$
($N\geq 2$). By employing techniques of Calculus of Variations we obtain the
multiplicity of solutions. The striking difference between our case and the
entire space case is also discussed.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,840 | Thermodynamic Stabilization of Precipitates through Interface Segregation: Chemical Effects | Precipitation hardening, which relies on a high density of intermetallic
precipitates, is a commonly utilized technique for strengthening structural
alloys. Structural alloys are commonly strengthened through a high density of
small size intermetallic precipitates. At high temperatures, however, the
precipitates coarsen to reduce the excess energy of the interface, resulting in
a significant reduction in the strengthening provided by the precipitates. In
certain ternary alloys, the secondary solute segregates to the interface and
results in the formation of a high density of nanosize precipitates that
provide enhanced strength and are resistant to coarsening. To understand the
chemical effects involved, and to identify such systems, we develop a
thermodynamic model using the framework of the regular nanocrystalline solution
model. For various global compositions, temperatures and thermodynamic
parameters, equilibrium configuration of Mg-Sn-Zn alloy is evaluated by
minimizing the Gibbs free energy function with respect to the region-specific
(bulk solid-solution, interface and precipitate) concentrations and sizes. The
results show that Mg$_2$Sn precipitates can be stabilized to nanoscale sizes
through Zn segregation to Mg/Mg$_2$Sn interface, and the precipitates can be
stabilized against coarsening at high-temperatures by providing a larger Zn
concentration in the system. Together with the inclusion of elastic strain
energy effects and the input of computationally informed interface
thermodynamic parameters in the future, the model is expected to provide a more
realistic prediction of segregation and precipitate stabilization in ternary
alloys of structural importance.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,841 | Wavelength Does Not Equal Pressure: Vertical Contribution Functions and their Implications for Mapping Hot Jupiters | Multi-band phase variations in principle allow us to infer the longitudinal
temperature distributions of planets as a function of height in their
atmospheres. For example, 3.6 micron emission originates from deeper layers of
the atmosphere than 4.5 micron due to greater water vapor absorption at the
longer wavelength. Since heat transport efficiency increases with pressure, we
expect thermal phase curves at 3.6 micron to exhibit smaller amplitudes and
greater phase offsets than at 4.5 micron; this trend is not observed. Of the
seven hot Jupiters with full-orbit phase curves at 3.6 and 4.5 micron, all have
greater phase amplitude at 3.6 micron than at 4.5 micron, while four of seven
exhibit a greater phase offset at 3.6 micron. We use a 3D
radiative-hydrodynamic model to calculate theoretical phase curves of HD
189733b, assuming thermo-chemical equilibrium. The model exhibits temperature,
pressure, and wavelength dependent opacity, primarily driven by carbon
chemistry: CO is energetically favored on the dayside, while CH4 is favored on
the cooler nightside. Infrared opacity therefore changes by orders of magnitude
between day and night, producing dramatic vertical shifts in the
wavelength-specific photospheres, which would complicate eclipse or phase
mapping with spectral data. The model predicts greater relative phase amplitude
and greater phase offset at 3.6 micron than at 4.5 micron, in agreement with
the data. Our model qualitatively explains the observed phase curves, but is in
tension with current thermo-chemical kinetics models that predict zonally
uniform atmospheric composition due to transport of CO from the hot regions of
the atmosphere.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,842 | Supervised Learning of Labeled Pointcloud Differences via Cover-Tree Entropy Reduction | We introduce a new algorithm, called CDER, for supervised machine learning
that merges the multi-scale geometric properties of Cover Trees with the
information-theoretic properties of entropy. CDER applies to a training set of
labeled pointclouds embedded in a common Euclidean space. If typical
pointclouds corresponding to distinct labels tend to differ at any scale in any
sub-region, CDER can identify these differences in (typically) linear time,
creating a set of distributional coordinates which act as a feature extraction
mechanism for supervised learning. We describe theoretical properties and
implementation details of CDER, and illustrate its benefits on several
synthetic examples.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,843 | Design Considerations for Proposed Fermilab Integrable RCS | Integrable optics is an innovation in particle accelerator design that
provides strong nonlinear focusing while avoiding parametric resonances. One
promising application of integrable optics is to overcome the traditional
limits on accelerator intensity imposed by betatron tune-spread and collective
instabilities. The efficacy of high-intensity integrable accelerators will be
undergo comprehensive testing over the next several years at the Fermilab
Integrable Optics Test Accelerator (IOTA) and the University of Maryland
Electron Ring (UMER). We propose an integrable Rapid-Cycling Synchrotron (iRCS)
as a replacement for the Fermilab Booster to achieve multi-MW beam power for
the Fermilab high-energy neutrino program. We provide a overview of the machine
parameters and discuss an approach to lattice optimization. Integrable optics
requires arcs with integer-pi phase advance followed by drifts with matched
beta functions. We provide an example integrable lattice with features of a
modern RCS - long dispersion-free drifts, low momentum compaction,
superperiodicity, chromaticity correction, separate-function magnets, and
bounded beta functions.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,844 | Consistent Estimation in General Sublinear Preferential Attachment Trees | We propose an empirical estimator of the preferential attachment function $f$
in the setting of general preferential attachment trees. Using a supercritical
continuous-time branching process framework, we prove the almost sure
consistency of the proposed estimator. We perform simulations to study the
empirical properties of our estimators.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,845 | Learned Watershed: End-to-End Learning of Seeded Segmentation | Learned boundary maps are known to outperform hand- crafted ones as a basis
for the watershed algorithm. We show, for the first time, how to train
watershed computation jointly with boundary map prediction. The estimator for
the merging priorities is cast as a neural network that is con- volutional
(over space) and recurrent (over iterations). The latter allows learning of
complex shape priors. The method gives the best known seeded segmentation
results on the CREMI segmentation challenge.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,846 | MAGIC Contributions to the 35th International Cosmic Ray Conference (ICRC2017) | MAGIC (Major Atmospheric Gamma Imaging Cherenkov) is a system of two 17 m
diameter, F/1.03 Imaging Atmospheric Cherenkov Telescopes (IACT). They are
dedicated to the observation of gamma rays from galactic and extragalactic
sources in the very high energy range (VHE, 30 GeV to 100 TeV). This submission
contains links to the proceedings for the 35th International Cosmic Ray
Conference (ICRC2017), held in Bexco, Busan, Korea from the 12th to the 17th of
July, 2017.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,847 | Efficient Hidden Vector Encryptions and Its Applications | Predicate encryption is a new paradigm of public key encryption that enables
searches on encrypted data. Using the predicate encryption, we can search
keywords or attributes on encrypted data without decrypting the ciphertexts. In
predicate encryption, a ciphertext is associated with attributes and a token
corresponds to a predicate. The token that corresponds to a predicate $f$ can
decrypt the ciphertext associated with attributes $x$ if and only if $f(x)=1$.
Hidden vector encryption (HVE) is a special kind of predicate encryption. In
this thesis, we consider the efficiency, the generality, and the security of
HVE schemes. The results of this thesis are described as follows.
The first results of this thesis are efficient HVE schemes where the token
consists of just four group elements and the decryption only requires four
bilinear map computations, independent of the number of attributes in the
ciphertext. The construction uses composite order bilinear groups and is
selectively secure under the well-known assumptions. The second results are
efficient HVE schemes that are secure under any kind of pairing types. To
achieve our goals, we proposed a general framework that converts HVE schemes
from composite order bilinear groups to prime order bilinear groups. Using the
framework, we convert the previous HVE schemes from composite order bilinear
groups to prime order bilinear groups. The third results are fully secure HVE
schemes with short tokens. Previous HVE schemes were proven to be secure only
in the selective security model where the capabilities of the adversaries are
severely restricted. Using the dual system encryption techniques, we construct
fully secure HVE schemes with match revealing property in composite order
groups.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,848 | The Galaxy Clustering Crisis in Abundance Matching | Galaxy clustering on small scales is significantly under-predicted by
sub-halo abundance matching (SHAM) models that populate (sub-)haloes with
galaxies based on peak halo mass, $M_{\rm peak}$. SHAM models based on the peak
maximum circular velocity, $V_{\rm peak}$, have had much better success. The
primary reason $M_{\rm peak}$ based models fail is the relatively low abundance
of satellite galaxies produced in these models compared to those based on
$V_{\rm peak}$. Despite success in predicting clustering, a simple $V_{\rm
peak}$ based SHAM model results in predictions for galaxy growth that are at
odds with observations. We evaluate three possible remedies that could "save"
mass-based SHAM: (1) SHAM models require a significant population of "orphan"
galaxies as a result of artificial disruption/merging of sub-haloes in modern
high resolution dark matter simulations; (2) satellites must grow significantly
after their accretion; and (3) stellar mass is significantly affected by halo
assembly history. No solution is entirely satisfactory. However, regardless of
the particulars, we show that popular SHAM models based on $M_{\rm peak}$
cannot be complete physical models as presented. Either $V_{\rm peak}$ truly is
a better predictor of stellar mass at $z\sim 0$ and it remains to be seen how
the correlation between stellar mass and $V_{\rm peak}$ comes about, or SHAM
models are missing vital component(s) that significantly affect galaxy
clustering.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,849 | Supplying Dark Energy from Scalar Field Dark Matter | We consider the hypothesis that dark matter and dark energy consists of
ultra-light self-interacting scalar particles. It is found that the
Klein-Gordon equation with only two free parameters (mass and self-coupling) on
a Schwarzschild background, at the galactic length-scales has the solution
which corresponds to Bose-Einstein condensate, behaving as dark matter, while
the constant solution at supra-galactic scales can explain dark energy.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,850 | GHz-Band Integrated Magnetic Inductors | The demand on mobile electronics to continue to shrink in size while increase
in efficiency drives the demand on the internal passive components to do the
same. Power amplifiers require inductors with small form factors, high quality
factors, and high operating frequency in the single-digit GHz range. This work
explores the use of magnetic materials to satisfy the needs of power amplifier
inductor applications. This paper discusses the optimization choices regarding
material selection, device design, and fabrication methodology. The inductors
achieved here present the best performance to date for an integrated magnetic
core inductor at high frequencies with a 1 nH inductance and peak quality
factor of 4 at ~3 GHz. Such compact inductors show potential for efficiently
meeting the need of mobile electronics in the future.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,851 | Acylindrical actions on projection complexes | We simplify the construction of projection complexes due to
Bestvina-Bromberg-Fujiwara. To do so, we introduce a sharper version of the
Behrstock inequality, and show that it can always be enforced. Furthermore, we
use the new setup to prove acylindricity results for the action on the
projection complexes. We also treat quasi-trees of metric spaces associated to
projection complexes, and prove an acylindricity criterion in that context as
well.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,852 | Introducing Geometric Algebra to Geometric Computing Software Developers: A Computational Thinking Approach | Designing software systems for Geometric Computing applications can be a
challenging task. Software engineers typically use software abstractions to
hide and manage the high complexity of such systems. Without the presence of a
unifying algebraic system to describe geometric models, the use of software
abstractions alone can result in many design and maintenance problems.
Geometric Algebra (GA) can be a universal abstract algebraic language for
software engineering geometric computing applications. Few sources, however,
provide enough information about GA-based software implementations targeting
the software engineering community. In particular, successfully introducing GA
to software engineers requires quite different approaches from introducing GA
to mathematicians or physicists. This article provides a high-level
introduction to the abstract concepts and algebraic representations behind the
elegant GA mathematical structure. The article focuses on the conceptual and
representational abstraction levels behind GA mathematics with sufficient
references for more details. In addition, the article strongly recommends
applying the methods of Computational Thinking in both introducing GA to
software engineers, and in using GA as a mathematical language for developing
Geometric Computing software systems.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,853 | Recovering Nonuniform Planted Partitions via Iterated Projection | In the planted partition problem, the $n$ vertices of a random graph are
partitioned into $k$ "clusters," and edges between vertices in the same cluster
and different clusters are included with constant probability $p$ and $q$,
respectively (where $0 \le q < p \le 1$). We give an efficient spectral
algorithm that recovers the clusters with high probability, provided that the
sizes of any two clusters are either very close or separated by $\geq
\Omega(\sqrt n)$. We also discuss a generalization of planted partition in
which the algorithm's input is not a random graph, but a random real symmetric
matrix with independent above-diagonal entries.
Our algorithm is an adaptation of a previous algorithm for the uniform case,
i.e., when all clusters are size $n / k \geq \Omega(\sqrt n)$. The original
algorithm recovers the clusters one by one via iterated projection: it
constructs the orthogonal projection operator onto the dominant $k$-dimensional
eigenspace of the random graph's adjacency matrix, uses it to recover one of
the clusters, then deletes it and recurses on the remaining vertices. We show
herein that a similar algorithm works in the nonuniform case.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,854 | Data-driven Analytics for Business Architectures: Proposed Use of Graph Theory | Business Architecture (BA) plays a significant role in helping organizations
understand enterprise structures and processes, and align them with strategic
objectives. However, traditional BAs are represented in fixed structure with
static model elements and fail to dynamically capture business insights based
on internal and external data. To solve this problem, this paper introduces the
graph theory into BAs with aim of building extensible data-driven analytics and
automatically generating business insights. We use IBM's Component Business
Model (CBM) as an example to illustrate various ways in which graph theory can
be leveraged for data-driven analytics, including what and how business
insights can be obtained. Future directions for applying graph theory to
business architecture analytics are discussed.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,855 | Beyond Sparsity: Tree Regularization of Deep Models for Interpretability | The lack of interpretability remains a key barrier to the adoption of deep
models in many applications. In this work, we explicitly regularize deep models
so human users might step through the process behind their predictions in
little time. Specifically, we train deep time-series models so their
class-probability predictions have high accuracy while being closely modeled by
decision trees with few nodes. Using intuitive toy examples as well as medical
tasks for treating sepsis and HIV, we demonstrate that this new tree
regularization yields models that are easier for humans to simulate than
simpler L1 or L2 penalties without sacrificing predictive power.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,856 | Phase-type distributions in population genetics | Probability modelling for DNA sequence evolution is well established and
provides a rich framework for understanding genetic variation between samples
of individuals from one or more populations. We show that both classical and
more recent models for coalescence (with or without recombination) can be
described in terms of the so-called phase-type theory, where complicated and
tedious calculations are circumvented by the use of matrices. The application
of phase-type theory consists of describing the stochastic model as a Markov
model by appropriately setting up a state space and calculating the
corresponding intensity and reward matrices. Formulae of interest are then
expressed in terms of these aforementioned matrices. We illustrate this by a
few examples calculating the mean, variance and even higher order moments of
the site frequency spectrum in the multiple merger coalescent models, and by
analysing the mean and variance for the number of segregating sites for
multiple samples in the two-locus ancestral recombination graph. We believe
that phase-type theory has great potential as a tool for analysing probability
models in population genetics. The compact matrix notation is useful for
clarification of current models, in particular their formal manipulation
(calculation), but also for further development or extensions.
| 0 | 0 | 0 | 1 | 1 | 0 |
20,857 | Robust Wald-type test in GLM with random design based on minimum density power divergence estimators | We consider the problem of robust inference under the important generalized
linear model (GLM) with stochastic covariates. We derive the properties of the
minimum density power divergence estimator of the parameters in GLM with random
design and used this estimator to propose a robust Wald-type test for testing
any general composite null hypothesis about the GLM. The asymptotic and
robustness properties of the proposed test are also examined for the GLM with
random design. Application of the proposed robust inference procedures to the
popular Poisson regression model for analyzing count data is discussed in
detail both theoretically and numerically with some interesting real data
examples.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,858 | Locally Smoothed Neural Networks | Convolutional Neural Networks (CNN) and the locally connected layer are
limited in capturing the importance and relations of different local receptive
fields, which are often crucial for tasks such as face verification, visual
question answering, and word sequence prediction. To tackle the issue, we
propose a novel locally smoothed neural network (LSNN) in this paper. The main
idea is to represent the weight matrix of the locally connected layer as the
product of the kernel and the smoother, where the kernel is shared over
different local receptive fields, and the smoother is for determining the
importance and relations of different local receptive fields. Specifically, a
multi-variate Gaussian function is utilized to generate the smoother, for
modeling the location relations among different local receptive fields.
Furthermore, the content information can also be leveraged by setting the mean
and precision of the Gaussian function according to the content. Experiments on
some variant of MNIST clearly show our advantages over CNN and locally
connected layer.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,859 | Asymptotic properties of a componentwise ARH(1) plug-in predictor | This paper presents new results on prediction of linear processes in function
spaces. The autoregressive Hilbertian process framework of order one (ARH(1)
process framework) is adopted. A componentwise estimator of the autocorrelation
operator is formulated, from the moment-based estimation of its diagonal
coefficients, with respect to the orthogonal eigenvectors of the
auto-covariance operator, which are assumed to be known. Mean-square
convergence to the theoretical autocorrelation operator, in the space of
Hilbert-Schmidt operators, is proved. Consistency then follows in that space.
For the associated ARH(1) plug-in predictor, mean absolute convergence to the
corresponding conditional expectation, in the considered Hilbert space, is
obtained. Hence, consistency in that space also holds. A simulation study is
undertaken to illustrate the finite-large sample behavior of the formulated
componentwise estimator and predictor. The performance of the presented
approach is compared with alternative approaches in the previous and current
ARH(1) framework literature, including the case of unknown eigenvectors.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,860 | The Effect of Population Control Policies on Societal Fragmentation | Population control policies are proposed and in some places employed as a
means towards curbing population growth. This paper is concerned with a
disturbing side-effect of such policies, namely, the potential risk of societal
fragmentation due to changes in the distribution of family sizes. This effect
is illustrated in some simple settings and demonstrated by simulation. In
addition, the dependence of societal fragmentation on family size distribution
is analyzed. In particular, it is shown that under the studied model, any
population control policy that disallows families of 3 or more children incurs
the possible risk of societal fragmentation.
| 1 | 1 | 0 | 0 | 0 | 0 |
20,861 | Optimizing Beam Transport in Rapidly Compressing Beams on the Neutralized Drift Compression Experiment - II | The Neutralized Drift Compression Experiment-II (NDCX-II) is an induction
linac that generates intense pulses of 1.2 MeV helium ions for heating matter
to extreme conditions. Here, we present recent results on optimizing beam
transport. The NDCX-II beamline includes a 1-meter-long drift section
downstream of the last transport solenoid, which is filled with
charge-neutralizing plasma that enables rapid longitudinal compression of an
intense ion beam against space-charge forces. The transport section on NDCX-II
consists of 28 solenoids. Finding optimal field settings for a group of
solenoids requires knowledge of the envelope parameters of the beam. Imaging
the beam on scintillator gives the radius of the beam, but the envelope angle
dr/dz is not measured directly. We demonstrate how the parameters of the beam
envelope (r, dr/dz, and emittance) can be reconstructed from a series of images
taken at varying B-field strengths of a solenoid upstream of the scintillator.
We use this technique to evaluate emittance at several points in the NDCX-II
beamline and for optimizing the trajectory of the beam at the entry of the
plasma-filled drift section.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,862 | Learn from Your Neighbor: Learning Multi-modal Mappings from Sparse Annotations | Many structured prediction problems (particularly in vision and language
domains) are ambiguous, with multiple outputs being correct for an input - e.g.
there are many ways of describing an image, multiple ways of translating a
sentence; however, exhaustively annotating the applicability of all possible
outputs is intractable due to exponentially large output spaces (e.g. all
English sentences). In practice, these problems are cast as multi-class
prediction, with the likelihood of only a sparse set of annotations being
maximized - unfortunately penalizing for placing beliefs on plausible but
unannotated outputs. We make and test the following hypothesis - for a given
input, the annotations of its neighbors may serve as an additional supervisory
signal. Specifically, we propose an objective that transfers supervision from
neighboring examples. We first study the properties of our developed method in
a controlled toy setup before reporting results on multi-label classification
and two image-grounded sequence modeling tasks - captioning and question
generation. We evaluate using standard task-specific metrics and measures of
output diversity, finding consistent improvements over standard maximum
likelihood training and other baselines.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,863 | Direct and indirect seismic inversion: interpretation of certain mathematical theorems | Quantitative methods are more familiar to most geophysicists with direct
inversion or indirect inversion. We will discuss seismic inversion in a high
level sense without getting into the actual algorithms. We will stay with
meta-equations and argue pros and cons based on certain mathematical theorems.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,864 | Kan's combinatorial spectra and their sheaves revisited | We define a right Cartan-Eilenberg structure on the category of Kan's
combinatorial spectra, and the category of sheaves of such spectra, assuming
some conditions. In both structures, we use the geometric concept of homotopy
equivalence as the strong equivalence. In the case of sheaves, we use local
equivalence as the weak equivalence. This paper is the first step in a
larger-scale program of investigating sheaves of spectra from a geometric
viewpoint.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,865 | Adiabatic Quantum Computing for Binary Clustering | Quantum computing for machine learning attracts increasing attention and
recent technological developments suggest that especially adiabatic quantum
computing may soon be of practical interest. In this paper, we therefore
consider this paradigm and discuss how to adopt it to the problem of binary
clustering. Numerical simulations demonstrate the feasibility of our approach
and illustrate how systems of qubits adiabatically evolve towards a solution.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,866 | Similarity Search Over Graphs Using Localized Spectral Analysis | This paper provides a new similarity detection algorithm. Given an input set
of multi-dimensional data points, where each data point is assumed to be
multi-dimensional, and an additional reference data point for similarity
finding, the algorithm uses kernel method that embeds the data points into a
low dimensional manifold. Unlike other kernel methods, which consider the
entire data for the embedding, our method selects a specific set of kernel
eigenvectors. The eigenvectors are chosen to separate between the data points
and the reference data point so that similar data points can be easily
identified as being distinct from most of the members in the dataset.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,867 | Reminiscences of Julian Schwinger: Late Harvard, Early UCLA Years (1968-1981) | These are reminiscences of my interactions with Julian Schwinger from 1968
through 1981 and beyond.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,868 | Site-resolved imaging of a bosonic Mott insulator using ytterbium atoms | We demonstrate site-resolved imaging of a strongly correlated quantum system
without relying on laser-cooling techniques during fluorescence imaging. We
observed the formation of Mott shells in the insulating regime and realized
thermometry on the atomic cloud. This work proves the feasibility of the
noncooled approach and opens the door to extending the detection technology to
new atomic species.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,869 | A unified continuum and variational multiscale formulation for fluids, solids, and fluid-structure interaction | We develop a unified continuum modeling framework for viscous fluids and
hyperelastic solids using the Gibbs free energy as the thermodynamic potential.
This framework naturally leads to a pressure primitive variable formulation for
the continuum body, which is well-behaved in both compressible and
incompressible regimes. Our derivation also provides a rational justification
of the isochoric-volumetric additive split of free energies in nonlinear
continuum mechanics. The variational multiscale analysis is performed for the
continuum model to construct a foundation for numerical discretization. We
first consider the continuum body instantiated as a hyperelastic material and
develop a variational multiscale formulation for the hyper-elastodynamic
problem. The generalized-alpha method is applied for temporal discretization. A
segregated algorithm for the nonlinear solver is designed and carefully
analyzed. Second, we apply the new formulation to construct a novel unified
formulation for fluid-solid coupled problems. The variational multiscale
formulation is utilized for spatial discretization in both fluid and solid
subdomains. The generalized-alpha method is applied for the whole continuum
body, and optimal high-frequency dissipation is achieved in both fluid and
solid subproblems. A new predictor multi-corrector algorithm is developed based
on the segregated algorithm to attain a good balance between robustness and
efficiency. The efficacy of the new formulations is examined in several
benchmark problems. The results indicate that the proposed modeling and
numerical methodologies constitute a promising technology for biomedical and
engineering applications, particularly those necessitating incompressible
models.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,870 | Time Assignment System and Its Performance aboard the Hitomi Satellite | Fast timing capability in X-ray observation of astrophysical objects is one
of the key properties for the ASTRO-H (Hitomi) mission. Absolute timing
accuracies of 350 micro second or 35 micro second are required to achieve
nominal scientific goals or to study fast variabilities of specific sources.
The satellite carries a GPS receiver to obtain accurate time information, which
is distributed from the central onboard computer through the large and complex
SpaceWire network. The details on the time system on the hardware and software
design are described. In the distribution of the time information, the
propagation delays and jitters affect the timing accuracy. Six other items
identified within the timing system will also contribute to absolute time
error. These error items have been measured and checked on ground to ensure the
time error budgets meet the mission requirements. The overall timing
performance in combination with hardware performance, software algorithm, and
the orbital determination accuracies, etc, under nominal conditions satisfies
the mission requirements of 35 micro second. This work demonstrates key points
for space-use instruments in hardware and software designs and calibration
measurements for fine timing accuracy on the order of microseconds for
mid-sized satellites using the SpaceWire (IEEE1355) network.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,871 | Common fixed points via $λ$-sequences in $G$-metric spaces | In this article, we use $\lambda$-sequences to derive common fixed points for
a family of self-mappings defined on a complete $G$-metric space. We imitate
some existing techniques in our proofs and show that the tools emlyed can be
used at a larger scale. These results generalise well known results in the
literature.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,872 | Bidirectional Nested Weighted Automata | Nested weighted automata (NWA) present a robust and convenient
automata-theoretic formalism for quantitative specifications. Previous works
have considered NWA that processed input words only in the forward direction.
It is natural to allow the automata to process input words backwards as well,
for example, to measure the maximal or average time between a response and the
preceding request. We therefore introduce and study bidirectional NWA that can
process input words in both directions. First, we show that bidirectional NWA
can express interesting quantitative properties that are not expressible by
forward-only NWA. Second, for the fundamental decision problems of emptiness
and universality, we establish decidability and complexity results for the new
framework which match the best-known results for the special case of
forward-only NWA. Thus, for NWA, the increased expressiveness of
bidirectionality is achieved at no additional computational complexity. This is
in stark contrast to the unweighted case, where bidirectional finite automata
are no more expressive but exponentially more succinct than their forward-only
counterparts.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,873 | Uniform Shapiro-Lopatinski conditions and boundary value problems on manifolds with bounded geometry | We study the regularity of the solutions of second order boundary value
problems on manifolds with boundary and bounded geometry. We first show that
the regularity property of a given boundary value problem $(P, C)$ is
equivalent to the uniform regularity of the natural family $(P_x, C_x)$ of
associated boundary value problems in local coordinates. We verify that this
property is satisfied for the Dirichlet boundary conditions and strongly
elliptic operators via a compactness argument. We then introduce a uniform
Shapiro-Lopatinski regularity condition, which is a modification of the
classical one, and we prove that it characterizes the boundary value problems
that satisfy the usual regularity property. We also show that the natural Robin
boundary conditions always satisfy the uniform Shapiro-Lopatinski regularity
condition, provided that our operator satisfies the strong Legendre condition.
This is achieved by proving that "well-posedness implies regularity" via a
modification of the classical "Nirenberg trick". When combining our regularity
results with the Poincaré inequality of (Ammann-Grosse-Nistor, preprint
2015), one obtains the usual well-posedness results for the classical boundary
value problems in the usual scale of Sobolev spaces, thus extending these
important, well-known theorems from smooth, bounded domains, to manifolds with
boundary and bounded geometry. As we show in several examples, these results do
not hold true anymore if one drops the bounded geometry assumption. We also
introduce a uniform Agmon condition and show that it is equivalent to the
coerciveness. Consequently, we prove a well-posedness result for parabolic
equations whose elliptic generator satisfies the uniform Agmon condition.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,874 | A Fourier-Chebyshev Spectral Method for Cavitation Computation in Nonlinear Elasticity | A Fourier-Chebyshev spectral method is proposed in this paper for solving the
cavitation problem in nonlinear elasticity. The interpolation error for the
cavitation solution is analyzed, the elastic energy error estimate for the
discrete cavitation solution is obtained, and the convergence of the method is
proved. An algorithm combined a gradient type method with a damped quasi-Newton
method is applied to solve the discretized nonlinear equilibrium equations.
Numerical experiments show that the Fourier-Chebyshev spectral method is
efficient and capable of producing accurate numerical cavitation solutions.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,875 | Learning Criticality in an Embodied Boltzmann Machine | Many biological and cognitive systems do not operate deep into one or other
regime of activity. Instead, they exploit critical surfaces poised at
transitions in their parameter space. The pervasiveness of criticality in
natural systems suggests that there may be general principles inducing this
behaviour. However, there is a lack of conceptual models explaining how
embodied agents propel themselves towards these critical points. In this paper,
we present a learning model driving an embodied Boltzmann Machine towards
critical behaviour by maximizing the heat capacity of the network. We test and
corroborate the model implementing an embodied agent in the mountain car
benchmark, controlled by a Boltzmann Machine that adjust its weights according
to the model. We find that the neural controller reaches a point of
criticality, which coincides with a transition point of the behaviour of the
agent between two regimes of behaviour, maximizing the synergistic information
between its sensors and the hidden and motor neurons. Finally, we discuss the
potential of our learning model to study the contribution of criticality to the
behaviour of embodied living systems in scenarios not necessarily constrained
by biological restrictions of the examples of criticality we find in nature.
| 1 | 1 | 0 | 0 | 0 | 0 |
20,876 | A contemporary look at Hermann Hankel's 1861 pioneering work on Lagrangian fluid dynamics | The present paper is a companion to the paper by Villone and Rampf (2017),
titled "Hermann Hankel's On the general theory of motion of fluids, an essay
including an English translation of the complete Preisschrift from 1861"
together with connected documents. Here we give a critical assessment of
Hankel's work, which covers many important aspects of fluid dynamics considered
from a Lagrangian-coordinates point of view: variational formulation in the
spirit of Hamilton for elastic (barotropic) fluids, transport (we would now say
Lie transport) of vorticity, the Lagrangian significance of Clebsch variables,
etc. Hankel's work is also put in the perspective of previous and future work.
Hence, the action spans about two centuries: from Lagrange's 1760-1761 Turin
paper on variational approaches to mechanics and fluid mechanics problems to
Arnold's 1966 founding paper on the geometrical/variational formulation of
incompressible flow. The 22-year old Hankel - who was to die 12 years later -
emerges as a highly innovative master of mathematical fluid dynamics, fully
deserving Riemann's assessment that his Preisschrift contains "all manner of
good things."
| 0 | 1 | 1 | 0 | 0 | 0 |
20,877 | A Dynamic Edge Exchangeable Model for Sparse Temporal Networks | We propose a dynamic edge exchangeable network model that can capture sparse
connections observed in real temporal networks, in contrast to existing models
which are dense. The model achieved superior link prediction accuracy on
multiple data sets when compared to a dynamic variant of the blockmodel, and is
able to extract interpretable time-varying community structures from the data.
In addition to sparsity, the model accounts for the effect of social influence
on vertices' future behaviours. Compared to the dynamic blockmodels, our model
has a smaller latent space. The compact latent space requires a smaller number
of parameters to be estimated in variational inference and results in a
computationally friendly inference algorithm.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,878 | Current-mode Memristor Crossbars for Neuromemristive Systems | Motivated by advantages of current-mode design, this brief contribution
explores the implementation of weight matrices in neuromemristive systems via
current-mode memristor crossbar circuits. After deriving theoretical results
for the range and distribution of weights in the current-mode design, it is
shown that any weight matrix based on voltage-mode crossbars can be mapped to a
current-mode crossbar if the voltage-mode weights are carefully bounded. Then,
a modified gradient descent rule is derived for the current-mode design that
can be used to perform backpropagation training. Behavioral simulations on the
MNIST dataset indicate that both voltage and current-mode designs are able to
achieve similar accuracy and have similar defect tolerance. However, analysis
of trained weight distributions reveals that current-mode and voltage-mode
designs may use different feature representations.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,879 | Incomplete Gauss sums modulo primes | We obtain a new bound for incomplete Gauss sums modulo primes. Our argument
falls under the framework of Vinogradov's method which we use to reduce the
problem under consideration to bounding the number of solutions to two distinct
systems of congruences. The first is related to Vinogradov's mean value
theorem, although the second does not appear to have been considered before.
Our bound improves on current results in the range $N\ge
q^{2k^{-1/2}+O(k^{-3/2})}$.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,880 | FAST Adaptive Smoothing and Thresholding for Improved Activation Detection in Low-Signal fMRI | Functional Magnetic Resonance Imaging is a noninvasive tool used to study
brain function. Detecting activation is challenged by many factors, and even
more so in low-signal scenarios that arise in the performance of high-level
cognitive tasks. We provide a fully automated and fast adaptive smoothing and
thresholding (FAST) algorithm that uses smoothing and extreme value theory on
correlated statistical parametric maps for thresholding. Performance on
experiments spanning a range of low-signal settings is very encouraging. The
methodology also performs well in a study to identify the cerebral regions that
perceive only-auditory-reliable and only-visual-reliable speech stimuli as well
as those that perceive one but not the other.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,881 | Improper multiferroicity and colossal dielectric constants in Bi$_{2}$CuO$_{4}$ | The layered cuprate Bi$_{2}$CuO$_{4}$ is investigated using magnetic,
dielectric and pyroelectric measurements. This system is observed to be an
improper multiferroic, with a robust ferroelectric state being established near
the magnetic transition. Magnetic and dielectric measurements indicate the
presence of a region above the antiferromagnetic Neel temperature with
concomitant polar and magnetic short range order. Bi$_{2}$CuO$_{4}$ is also
seen to exhibit colossal dielectric constants at higher temperatures with
clearly distinguishable grain and grain boundary contributions, both of which
exhibit non-Debye relaxation.
| 0 | 1 | 0 | 0 | 0 | 0 |
20,882 | Suppression of plasma echoes and Landau damping in Sobolev spaces by weak collisions in a Vlasov-Fokker-Planck equation | In this paper, we study Landau damping in the weakly collisional limit of a
Vlasov-Fokker-Planck equation with nonlinear collisions in the phase-space
$(x,v) \in \mathbb T_x^n \times \mathbb R^n_v$. The goal is four-fold: (A) to
understand how collisions suppress plasma echoes and enable Landau damping in
agreement with linearized theory in Sobolev spaces, (B) to understand how phase
mixing accelerates collisional relaxation, (C) to understand better how the
plasma returns to global equilibrium during Landau damping, and (D) to rule out
that collision-driven nonlinear instabilities dominate. We give an estimate for
the scaling law between Knudsen number and the maximal size of the perturbation
necessary for linear theory to be accurate in Sobolev regularity. We conjecture
this scaling to be sharp (up to logarithmic corrections) due to potential
nonlinear echoes in the collisionless model.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,883 | Deep Learning for Semantic Segmentation on Minimal Hardware | Deep learning has revolutionised many fields, but it is still challenging to
transfer its success to small mobile robots with minimal hardware.
Specifically, some work has been done to this effect in the RoboCup humanoid
football domain, but results that are performant and efficient and still
generally applicable outside of this domain are lacking. We propose an approach
conceptually different from those taken previously. It is based on semantic
segmentation and does achieve these desired properties. In detail, it is being
able to process full VGA images in real-time on a low-power mobile processor.
It can further handle multiple image dimensions without retraining, it does not
require specific domain knowledge for achieving a high frame rate and it is
applicable on a minimal mobile hardware.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,884 | Overpartition $M2$-rank differences, class number relations, and vector-valued mock Eisenstein series | We prove that the generating function of overpartition $M2$-rank differences
is, up to coefficient signs, a component of the vector-valued mock Eisenstein
series attached to a certain quadratic form. We use this to compute analogs of
the class number relations for $M2$-rank differences. As applications we split
the Kronecker-Hurwitz relation into its "even" and "odd" parts and calculate
sums over Hurwitz class numbers of the form $\sum_{r \in \mathbb{Z}} H(n -
2r^2)$.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,885 | Communication-Efficient and Decentralized Multi-Task Boosting while Learning the Collaboration Graph | We study the decentralized machine learning scenario where many users
collaborate to learn personalized models based on (i) their local datasets and
(ii) a similarity graph over the users' learning tasks. Our approach trains
nonlinear classifiers in a multi-task boosting manner without exchanging
personal data and with low communication costs. When background knowledge about
task similarities is not available, we propose to jointly learn the
personalized models and a sparse collaboration graph through an alternating
optimization procedure. We analyze the convergence rate, memory consumption and
communication complexity of our decentralized algorithms, and demonstrate the
benefits of our approach compared to competing techniques on synthetic and real
datasets.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,886 | On the well-posedness of SPDEs with singular drift in divergence form | We prove existence and uniqueness of strong solutions for a class of
second-order stochastic PDEs with multiplicative Wiener noise and drift of the
form $\operatorname{div} \gamma(\nabla \cdot)$, where $\gamma$ is a maximal
monotone graph in $\mathbb{R}^n \times \mathbb{R}^n$ obtained as the
subdifferential of a convex function satisfying very mild assumptions on its
behavior at infinity. The well-posedness result complements the corresponding
one in our recent work arXiv:1612.08260 where, under the additional assumption
that $\gamma$ is single-valued, a solution with better integrability and
regularity properties is constructed. The proof given here, however, is
self-contained.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,887 | Trace norm regularization and faster inference for embedded speech recognition RNNs | We propose and evaluate new techniques for compressing and speeding up dense
matrix multiplications as found in the fully connected and recurrent layers of
neural networks for embedded large vocabulary continuous speech recognition
(LVCSR). For compression, we introduce and study a trace norm regularization
technique for training low rank factored versions of matrix multiplications.
Compared to standard low rank training, we show that our method leads to good
accuracy versus number of parameter trade-offs and can be used to speed up
training of large models. For speedup, we enable faster inference on ARM
processors through new open sourced kernels optimized for small batch sizes,
resulting in 3x to 7x speed ups over the widely used gemmlowp library. Beyond
LVCSR, we expect our techniques and kernels to be more generally applicable to
embedded neural networks with large fully connected or recurrent layers.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,888 | Scalable and Robust Sparse Subspace Clustering Using Randomized Clustering and Multilayer Graphs | Sparse subspace clustering (SSC) is one of the current state-of-the-art
methods for partitioning data points into the union of subspaces, with strong
theoretical guarantees. However, it is not practical for large data sets as it
requires solving a LASSO problem for each data point, where the number of
variables in each LASSO problem is the number of data points. To improve the
scalability of SSC, we propose to select a few sets of anchor points using a
randomized hierarchical clustering method, and, for each set of anchor points,
solve the LASSO problems for each data point allowing only anchor points to
have a non-zero weight (this reduces drastically the number of variables). This
generates a multilayer graph where each layer corresponds to a different set of
anchor points. Using the Grassmann manifold of orthogonal matrices, the shared
connectivity among the layers is summarized within a single subspace. Finally,
we use $k$-means clustering within that subspace to cluster the data points,
similarly as done by spectral clustering in SSC. We show on both synthetic and
real-world data sets that the proposed method not only allows SSC to scale to
large-scale data sets, but that it is also much more robust as it performs
significantly better on noisy data and on data with close susbspaces and
outliers, while it is not prone to oversegmentation.
| 0 | 0 | 0 | 1 | 0 | 0 |
20,889 | Deep Health Care Text Classification | Health related social media mining is a valuable apparatus for the early
recognition of the diverse antagonistic medicinal conditions. Mostly, the
existing methods are based on machine learning with knowledge-based learning.
This working note presents the Recurrent neural network (RNN) and Long
short-term memory (LSTM) based embedding for automatic health text
classification in the social media mining. For each task, two systems are built
and that classify the tweet at the tweet level. RNN and LSTM are used for
extracting features and non-linear activation function at the last layer
facilitates to distinguish the tweets of different categories. The experiments
are conducted on 2nd Social Media Mining for Health Applications Shared Task at
AMIA 2017. The experiment results are considerable; however the proposed method
is appropriate for the health text classification. This is primarily due to the
reason that, it doesn't rely on any feature engineering mechanisms.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,890 | Study of cost functionals for ptychographic phase retrieval to improve the robustness against noise, and a proposal for another noise-robust ptychographic phase retrieval scheme | Recently, efforts have been made to improve ptychography phase retrieval
algorithms so that they are more robust against noise. Often the algorithm is
adapted by changing the cost functional that needs to be minimized. In
particular, it has been suggested that the cost functional should be obtained
using a maximum-likelihood approach that takes the noise statistics into
account. Here, we consider the different choices of cost functional, and to how
they affect the reconstruction results. We find that seemingly the only
consistently reliable way to improve reconstruction results in the presence of
noise is to reduce the step size of the update function. In addition, a
noise-robust ptychographic reconstruction method has been proposed that relies
on adapting the intensity constraints
| 1 | 1 | 0 | 0 | 0 | 0 |
20,891 | Predicting Native Language from Gaze | A fundamental question in language learning concerns the role of a speaker's
first language in second language acquisition. We present a novel methodology
for studying this question: analysis of eye-movement patterns in second
language reading of free-form text. Using this methodology, we demonstrate for
the first time that the native language of English learners can be predicted
from their gaze fixations when reading English. We provide analysis of
classifier uncertainty and learned features, which indicates that differences
in English reading are likely to be rooted in linguistic divergences across
native languages. The presented framework complements production studies and
offers new ground for advancing research on multilingualism.
| 1 | 0 | 0 | 0 | 0 | 0 |
20,892 | A form of Schwarz's lemma and a bound for the Kobayashi metric on convex domains | We present a form of Schwarz's lemma for holomorphic maps between convex
domains $D_1$ and $D_2$. This result provides a lower bound on the distance
between the images of relatively compact subsets of $D_1$ and the boundary of
$D_2$. This is a natural improvement of an old estimate by Bernal-González
that takes into account the geometry of $\partial{D_1}$. We also provide a new
estimate for the Kobayashi metric on bounded convex domains.
| 0 | 0 | 1 | 0 | 0 | 0 |
20,893 | The COM-negative binomial distribution: modeling overdispersion and ultrahigh zero-inflated count data | In this paper, we focus on the COM-type negative binomial distribution with
three parameters, which belongs to COM-type $(a,b,0)$ class distributions and
family of equilibrium distributions of arbitrary birth-death process. Besides,
we show abundant distributional properties such as overdispersion and
underdispersion, log-concavity, log-convexity (infinite divisibility), pseudo
compound Poisson, stochastic ordering and asymptotic approximation. Some
characterizations including sum of equicorrelated geometrically distributed
random variables, conditional distribution, limit distribution of COM-negative
hypergeometric distribution, and Stein's identity are given for theoretical
properties. COM-negative binomial distribution was applied to overdispersion
and ultrahigh zero-inflated data sets. With the aid of ratio regression, we
employ maximum likelihood method to estimate the parameters and the
goodness-of-fit are evaluated by the discrete Kolmogorov-Smirnov test.
| 0 | 0 | 1 | 1 | 0 | 0 |
20,894 | Efficient Contextual Bandits in Non-stationary Worlds | Most contextual bandit algorithms minimize regret against the best fixed
policy, a questionable benchmark for non-stationary environments that are
ubiquitous in applications. In this work, we develop several efficient
contextual bandit algorithms for non-stationary environments by equipping
existing methods for i.i.d. problems with sophisticated statistical tests so as
to dynamically adapt to a change in distribution.
We analyze various standard notions of regret suited to non-stationary
environments for these algorithms, including interval regret, switching regret,
and dynamic regret. When competing with the best policy at each time, one of
our algorithms achieves regret $\mathcal{O}(\sqrt{ST})$ if there are $T$ rounds
with $S$ stationary periods, or more generally
$\mathcal{O}(\Delta^{1/3}T^{2/3})$ where $\Delta$ is some non-stationarity
measure. These results almost match the optimal guarantees achieved by an
inefficient baseline that is a variant of the classic Exp4 algorithm. The
dynamic regret result is also the first one for efficient and fully adversarial
contextual bandit.
Furthermore, while the results above require tuning a parameter based on the
unknown quantity $S$ or $\Delta$, we also develop a parameter free algorithm
achieving regret $\min\{S^{1/4}T^{3/4}, \Delta^{1/5}T^{4/5}\}$. This improves
and generalizes the best existing result $\Delta^{0.18}T^{0.82}$ by Karnin and
Anava (2016) which only holds for the two-armed bandit problem.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,895 | Search Engine Drives the Evolution of Social Networks | The search engine is tightly coupled with social networks and is primarily
designed for users to acquire interested information. Specifically, the search
engine assists the information dissemination for social networks, i.e.,
enabling users to access interested contents with keywords-searching and
promoting the process of contents-transferring from the source users directly
to potential interested users. Accompanying such processes, the social network
evolves as new links emerge between users with common interests. However, there
is no clear understanding of such a "chicken-and-egg" problem, namely, new
links encourage more social interactions, and vice versa. In this paper, we aim
to quantitatively characterize the social network evolution phenomenon driven
by a search engine. First, we propose a search network model for social network
evolution. Second, we adopt two performance metrics, namely, degree
distribution and network diameter. Theoretically, we prove that the degree
distribution follows an intensified power-law, and the network diameter
shrinks. Third, we quantitatively show that the search engine accelerates the
rumor propagation in social networks. Finally, based on four real-world data
sets (i.e., CDBLP, Facebook, Weibo Tweets, P2P), we verify our theoretical
findings. Furthermore, we find that the search engine dramatically increases
the speed of rumor propagation.
| 1 | 1 | 0 | 0 | 0 | 0 |
20,896 | End-to-End Learning of Semantic Grasping | We consider the task of semantic robotic grasping, in which a robot picks up
an object of a user-specified class using only monocular images. Inspired by
the two-stream hypothesis of visual reasoning, we present a semantic grasping
framework that learns object detection, classification, and grasp planning in
an end-to-end fashion. A "ventral stream" recognizes object class while a
"dorsal stream" simultaneously interprets the geometric relationships necessary
to execute successful grasps. We leverage the autonomous data collection
capabilities of robots to obtain a large self-supervised dataset for training
the dorsal stream, and use semi-supervised label propagation to train the
ventral stream with only a modest amount of human supervision. We
experimentally show that our approach improves upon grasping systems whose
components are not learned end-to-end, including a baseline method that uses
bounding box detection. Furthermore, we show that jointly training our model
with auxiliary data consisting of non-semantic grasping data, as well as
semantically labeled images without grasp actions, has the potential to
substantially improve semantic grasping performance.
| 1 | 0 | 0 | 1 | 0 | 0 |
20,897 | Von Neumann Regular Cellular Automata | For any group $G$ and any set $A$, a cellular automaton (CA) is a
transformation of the configuration space $A^G$ defined via a finite memory set
and a local function. Let $\text{CA}(G;A)$ be the monoid of all CA over $A^G$.
In this paper, we investigate a generalisation of the inverse of a CA from the
semigroup-theoretic perspective. An element $\tau \in \text{CA}(G;A)$ is von
Neumann regular (or simply regular) if there exists $\sigma \in \text{CA}(G;A)$
such that $\tau \circ \sigma \circ \tau = \tau$ and $\sigma \circ \tau \circ
\sigma = \sigma$, where $\circ$ is the composition of functions. Such an
element $\sigma$ is called a generalised inverse of $\tau$. The monoid
$\text{CA}(G;A)$ itself is regular if all its elements are regular. We
establish that $\text{CA}(G;A)$ is regular if and only if $\vert G \vert = 1$
or $\vert A \vert = 1$, and we characterise all regular elements in
$\text{CA}(G;A)$ when $G$ and $A$ are both finite. Furthermore, we study
regular linear CA when $A= V$ is a vector space over a field $\mathbb{F}$; in
particular, we show that every regular linear CA is invertible when $G$ is
torsion-free elementary amenable (e.g. when $G=\mathbb{Z}^d, \ d \in
\mathbb{N}$) and $V=\mathbb{F}$, and that every linear CA is regular when $V$
is finite-dimensional and $G$ is locally finite with $\text{Char}(\mathbb{F})
\nmid o(g)$ for all $g \in G$.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,898 | A Distance Between Filtered Spaces Via Tripods | We present a simplified treatment of stability of filtrations on finite
spaces. Interestingly, we can lift the stability result for combinatorial
filtrations from [CSEM06] to the case when two filtrations live on different
spaces without directly invoking the concept of interleaving. We then prove
that this distance is intrinsic by constructing explicit geodesics between any
pair of filtered spaces. Finally we use this construction to obtain a
strengthening of the stability result.
| 1 | 0 | 1 | 0 | 0 | 0 |
20,899 | Robust Shape Estimation for 3D Deformable Object Manipulation | Existing shape estimation methods for deformable object manipulation suffer
from the drawbacks of being off-line, model dependent, noise-sensitive or
occlusion-sensitive, and thus are not appropriate for manipulation tasks
requiring high precision. In this paper, we present a real-time shape
estimation approach for autonomous robotic manipulation of 3D deformable
objects. Our method fulfills all the requirements necessary for the
high-quality deformable object manipulation in terms of being real-time,
model-free and robust to noise and occlusion. These advantages are accomplished
using a joint tracking and reconstruction framework, in which we track the
object deformation by aligning a reference shape model with the stream input
from the RGB-D camera, and simultaneously upgrade the reference shape model
according to the newly captured RGB-D data. We have evaluated the quality and
robustness of our real-time shape estimation pipeline on a set of deformable
manipulation tasks implemented on physical robots. Videos are available at
this https URL
| 1 | 0 | 0 | 0 | 0 | 0 |
20,900 | Beyond Winning and Losing: Modeling Human Motivations and Behaviors Using Inverse Reinforcement Learning | In recent years, reinforcement learning (RL) methods have been applied to
model gameplay with great success, achieving super-human performance in various
environments, such as Atari, Go, and Poker. However, those studies mostly focus
on winning the game and have largely ignored the rich and complex human
motivations, which are essential for understanding different players' diverse
behaviors. In this paper, we present a novel method called Multi-Motivation
Behavior Modeling (MMBM) that takes the multifaceted human motivations into
consideration and models the underlying value structure of the players using
inverse RL. Our approach does not require the access to the dynamic of the
system, making it feasible to model complex interactive environments such as
massively multiplayer online games. MMBM is tested on the World of Warcraft
Avatar History dataset, which recorded over 70,000 users' gameplay spanning
three years period. Our model reveals the significant difference of value
structures among different player groups. Using the results of motivation
modeling, we also predict and explain their diverse gameplay behaviors and
provide a quantitative assessment of how the redesign of the game environment
impacts players' behaviors.
| 0 | 0 | 0 | 1 | 0 | 0 |
Subsets and Splits