text
stringlengths 138
2.38k
| labels
sequencelengths 6
6
| Predictions
sequencelengths 1
3
|
|---|---|---|
Title: Cubical-like geometry of quasi-median graphs and applications to geometric group theory,
Abstract: The class of quasi-median graphs is a generalisation of median graphs, or
equivalently of CAT(0) cube complexes. The purpose of this thesis is to
introduce these graphs in geometric group theory. In the first part of our
work, we extend the definition of hyperplanes from CAT(0) cube complexes, and
we show that the geometry of a quasi-median graph essentially reduces to the
combinatorics of its hyperplanes. In the second part, we exploit the specific
structure of the hyperplanes to state combination results. The main idea is
that if a group acts in a suitable way on a quasi-median graph so that
clique-stabilisers satisfy some non-positively curved property $\mathcal{P}$,
then the whole group must satisfy $\mathcal{P}$ as well. The properties we are
interested in are mainly (relative) hyperbolicity, (equivariant)
$\ell^p$-compressions, CAT(0)-ness and cubicality. In the third part, we apply
our general criteria to several classes of groups, including graph products,
Guba and Sapir's diagram products, some wreath products, and some graphs of
groups. Graph products are our most natural examples, where the link between
the group and its quasi-median graph is particularly strong and explicit; in
particular, we are able to determine precisely when a graph product is
relatively hyperbolic. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Doubly Nested Network for Resource-Efficient Inference,
Abstract: We propose doubly nested network(DNNet) where all neurons represent their own
sub-models that solve the same task. Every sub-model is nested both layer-wise
and channel-wise. While nesting sub-models layer-wise is straight-forward with
deep-supervision as proposed in \cite{xie2015holistically}, channel-wise
nesting has not been explored in the literature to our best knowledge.
Channel-wise nesting is non-trivial as neurons between consecutive layers are
all connected to each other. In this work, we introduce a technique to solve
this problem by sorting channels topologically and connecting neurons
accordingly. For the purpose, channel-causal convolutions are used. Slicing
doubly nested network gives a working sub-network. The most notable application
of our proposed network structure with slicing operation is resource-efficient
inference. At test time, computing resources such as time and memory available
for running the prediction algorithm can significantly vary across devices and
applications. Given a budget constraint, we can slice the network accordingly
and use a sub-model for inference within budget, requiring no additional
computation such as training or fine-tuning after deployment. We demonstrate
the effectiveness of our approach in several practical scenarios of utilizing
available resource efficiently. | [
0,
0,
0,
1,
0,
0
] | [
"Computer Science"
] |
Title: Structural and bonding character of potassium-doped p-terphenyl superconductors,
Abstract: Recently, there is a series of reports by Wang et al. on the
superconductivity in K-doped p-terphenyl (KxC18H14) with the transition
temperatures range from 7 to 123 Kelvin. Identifying the structural and bonding
character is the key to understand the superconducting phases and the related
properties. Therefore we carried out an extensive study on the crystal
structures with different doping levels and investigate the thermodynamic
stability, structural, electronic, and magnetic properties by the
first-principles calculations. Our calculated structures capture most features
of the experimentally observed X-ray diffraction patterns. The K doping
concentration is constrained to within the range of 2 and 3. The obtained
formation energy indicates that the system at x = 2.5 is more stable. The
strong ionic bonding interaction is found in between K atoms and organic
molecules. The charge transfer accounts for the metallic feature of the doped
materials. For a small amount of charge transferred, the tilting force between
the two successive benzenes drives the system to stabilize at the
antiferromagnetic ground state, while the system exhibits non-magnetic behavior
with increasing charge transfer. The multiformity of band structures near the
Fermi level indicates that the driving force for superconductivity is
complicated. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Influence of the Forward Difference Scheme for the Time Derivative on the Stability of Wave Equation Numerical Solution,
Abstract: Research on numerical stability of difference equations has been quite
intensive in the past century. The choice of difference schemes for the
derivative terms in these equations contributes to a wide range of the
stability analysis issues - one of which is how a chosen scheme may directly or
indirectly contribute to such stability. In the present paper, how far the
forward difference scheme for the time derivative in the wave equation
influences the stability of the equation numerical solution, is particularly
investigated. The stability analysis of the corresponding difference equation
involving four schemes, namely Lax's, central, forward, and rearward
differences, were carried out, and the resulting stability criteria were
compared. The results indicate that the instability of the solution of wave
equation is not always due to the forward difference scheme for the time
derivative. Rather, it is shown in this paper that the stability criterion is
still possible when the spatial derivative is represented by an appropriate
difference scheme. This sheds light on the degree of applicability of a
difference scheme for a hyperbolic equation. | [
1,
0,
0,
0,
0,
0
] | [
"Mathematics",
"Physics",
"Computer Science"
] |
Title: A Framework for Evaluating Model-Driven Self-adaptive Software Systems,
Abstract: In the last few years, Model Driven Development (MDD), Component-based
Software Development (CBSD), and context-oriented software have become
interesting alternatives for the design and construction of self-adaptive
software systems. In general, the ultimate goal of these technologies is to be
able to reduce development costs and effort, while improving the modularity,
flexibility, adaptability, and reliability of software systems. An analysis of
these technologies shows them all to include the principle of the separation of
concerns, and their further integration is a key factor to obtaining
high-quality and self-adaptable software systems. Each technology identifies
different concerns and deals with them separately in order to specify the
design of the self-adaptive applications, and, at the same time, support
software with adaptability and context-awareness. This research studies the
development methodologies that employ the principles of model-driven
development in building self-adaptive software systems. To this aim, this
article proposes an evaluation framework for analysing and evaluating the
features of model-driven approaches and their ability to support software with
self-adaptability and dependability in highly dynamic contextual environment.
Such evaluation framework can facilitate the software developers on selecting a
development methodology that suits their software requirements and reduces the
development effort of building self-adaptive software systems. This study
highlights the major drawbacks of the propped model-driven approaches in the
related works, and emphasise on considering the volatile aspects of
self-adaptive software in the analysis, design and implementation phases of the
development methodologies. In addition, we argue that the development
methodologies should leave the selection of modelling languages and modelling
tools to the software developers. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: Second order necessary and sufficient optimality conditions for singular solutions of partially-affine control problems,
Abstract: In this article we study optimal control problems for systems that are affine
with respect to some of the control variables and nonlinear in relation to the
others. We consider finitely many equality and inequality constraints on the
initial and final values of the state. We investigate singular optimal
solutions for this class of problems, for which we obtain second order
necessary and sufficient conditions for weak optimality in integral form. We
also derive Goh pointwise necessary optimality conditions. We show an example
to illustrate the results. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Phase diagram of a generalized off-diagonal Aubry-André model with p-wave pairing,
Abstract: Off-diagonal Aubry-André (AA) model has recently attracted a great deal
of attention as they provide condensed matter realization of topological
phases. We numerically study a generalized off-diagonal AA model with p-wave
superfluid pairing in the presence of both commensurate and incommensurate
hopping modulations. The phase diagram as functions of the modulation strength
of incommensurate hopping and the strength of the p-wave pairing is obtained by
using the multifractal analysis. We show that with the appearance of the p-wave
pairing, the system exhibits mobility-edge phases and critical phases with
various number of topologically protected zero-energy modes. Predicted
topological nature of these exotic phases can be realized in a cold atomic
system of incommensurate bichromatic optical lattice with induced p-wave
superfluid pairing by using a Raman laser in proximity to a molecular
Bose-Einstein condensation. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Around Average Behavior: 3-lambda Network Model,
Abstract: The analysis of networks affects the research of many real phenomena. The
complex network structure can be viewed as a network's state at the time of the
analysis or as a result of the process through which the network arises.
Research activities focus on both and, thanks to them, we know not only many
measurable properties of networks but also the essence of some phenomena that
occur during the evolution of networks. One typical research area is the
analysis of co-authorship networks and their evolution. In our paper, the
analysis of one real-world co-authorship network and inspiration from existing
models form the basis of the hypothesis from which we derive new 3-lambda
network model. This hypothesis works with the assumption that regular behavior
of nodes revolves around an average. However, some anomalies may occur. The
3-lambda model is stochastic and uses the three parameters associated with the
average behavior of the nodes. The growth of the network based on this model
assumes that one step of the growth is an interaction in which both new and
existing nodes are participating. In the paper we present the results of the
analysis of a co-authorship network and formulate a hypothesis and a model
based on this hypothesis. Later in the paper, we examine the outputs from the
network generator based on the 3-lambda model and show that generated networks
have characteristics known from the environment of real-world networks. | [
1,
1,
0,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: Informed Asymptotically Optimal Anytime Search,
Abstract: Path planning in robotics often requires finding high-quality solutions to
continuously valued and/or high-dimensional problems. These problems are
challenging and most planning algorithms instead solve simplified
approximations. Popular approximations include graphs and random samples, as
respectively used by informed graph-based searches and anytime sampling-based
planners. Informed graph-based searches, such as A*, traditionally use
heuristics to search a priori graphs in order of potential solution quality.
This makes their search efficient but leaves their performance dependent on the
chosen approximation. If its resolution is too low then they may not find a
(suitable) solution but if it is too high then they may take a prohibitively
long time to do so. Anytime sampling-based planners, such as RRT*,
traditionally use random sampling to approximate the problem domain
incrementally. This allows them to increase resolution until a suitable
solution is found but makes their search dependent on the order of
approximation. Arbitrary sequences of random samples expand the approximation
in every direction and fill the problem domain but may be prohibitively
inefficient at containing a solution. This paper unifies and extends these two
approaches to develop Batch Informed Trees (BIT*), an informed, anytime
sampling-based planner. BIT* solves continuous path planning problems
efficiently by using sampling and heuristics to alternately approximate and
search the problem domain. Its search is ordered by potential solution quality,
as in A*, and its approximation improves indefinitely with additional
computational time, as in RRT*. It is shown analytically to be almost-surely
asymptotically optimal and experimentally to outperform existing sampling-based
planners, especially on high-dimensional planning problems. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: Analysis of error control in large scale two-stage multiple hypothesis testing,
Abstract: When dealing with the problem of simultaneously testing a large number of
null hypotheses, a natural testing strategy is to first reduce the number of
tested hypotheses by some selection (screening or filtering) process, and then
to simultaneously test the selected hypotheses. The main advantage of this
strategy is to greatly reduce the severe effect of high dimensions. However,
the first screening or selection stage must be properly accounted for in order
to maintain some type of error control. In this paper, we will introduce a
selection rule based on a selection statistic that is independent of the test
statistic when the tested hypothesis is true. Combining this selection rule and
the conventional Bonferroni procedure, we can develop a powerful and valid
two-stage procedure. The introduced procedure has several nice properties: (i)
it completely removes the selection effect; (ii) it reduces the multiplicity
effect; (iii) it does not "waste" data while carrying out both selection and
testing. Asymptotic power analysis and simulation studies illustrate that this
proposed method can provide higher power compared to usual multiple testing
methods while controlling the Type 1 error rate. Optimal selection thresholds
are also derived based on our asymptotic analysis. | [
0,
0,
1,
1,
0,
0
] | [
"Statistics",
"Mathematics"
] |
Title: XES Tensorflow - Process Prediction using the Tensorflow Deep-Learning Framework,
Abstract: Predicting the next activity of a running process is an important aspect of
process management. Recently, artificial neural networks, so called
deep-learning approaches, have been proposed to address this challenge. This
demo paper describes a software application that applies the Tensorflow
deep-learning framework to process prediction. The software application reads
industry-standard XES files for training and presents the user with an
easy-to-use graphical user interface for both training and prediction. The
system provides several improvements over earlier work. This demo paper focuses
on the software implementation and describes the architecture and user
interface. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: An initial-boundary value problem for the integrable spin-1 Gross-Pitaevskii equations with a 4x4 Lax pair on the half-line,
Abstract: We investigate the initial-boundary value problem for the integrable spin-1
Gross-Pitaevskii (GP) equations with a 4x4 Lax pair on the half-line. The
solution of this system can be obtained in terms of the solution of a 4x4
matrix Riemann-Hilbert (RH) problem formulated in the complex k-plane. The
relevant jump matrices of the RH problem can be explicitly found using the two
spectral functions s(k) and S(k), which can be defined by the initial data, the
Dirichlet-Neumann boundary data at x=0. The global relation is established
between the two dependent spectral functions. The general mappings between
Dirichlet and Neumann boundary values are analyzed in terms of the global
relation. | [
0,
1,
1,
0,
0,
0
] | [
"Mathematics",
"Physics"
] |
Title: 4-DoF Tracking for Robot Fine Manipulation Tasks,
Abstract: This paper presents two visual trackers from the different paradigms of
learning and registration based tracking and evaluates their application in
image based visual servoing. They can track object motion with four degrees of
freedom (DoF) which, as we will show here, is sufficient for many fine
manipulation tasks. One of these trackers is a newly developed learning based
tracker that relies on learning discriminative correlation filters while the
other is a refinement of a recent 8 DoF RANSAC based tracker adapted with a new
appearance model for tracking 4 DoF motion.
Both trackers are shown to provide superior performance to several state of
the art trackers on an existing dataset for manipulation tasks. Further, a new
dataset with challenging sequences for fine manipulation tasks captured from
robot mounted eye-in-hand (EIH) cameras is also presented. These sequences have
a variety of challenges encountered during real tasks including jittery camera
movement, motion blur, drastic scale changes and partial occlusions.
Quantitative and qualitative results on these sequences are used to show that
these two trackers are robust to failures while providing high precision that
makes them suitable for such fine manipulation tasks. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: New ideas for tests of Lorentz invariance with atomic systems,
Abstract: We describe a broadly applicable experimental proposal to search for the
violation of local Lorentz invariance (LLI) with atomic systems. The new scheme
uses dynamic decoupling and can be implemented in current atomic clocks
experiments, both with single ions and arrays of neutral atoms. Moreover, the
scheme can be performed on systems with no optical transitions, and therefore
it is also applicable to highly charged ions which exhibit particularly high
sensitivity to Lorentz invariance violation. We show the results of an
experiment measuring the expected signal of this proposal using a two-ion
crystal of $^{88}$Sr$^+$ ions. We also carry out a systematic study of the
sensitivity of highly charged ions to LLI to identify the best candidates for
the LLI tests. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Small sets in dense pairs,
Abstract: Let $\widetilde{\mathcal M}=\langle \mathcal M, P\rangle$ be an expansion of
an o-minimal structure $\mathcal M$ by a dense set $P\subseteq M$, such that
three tameness conditions hold. We prove that the induced structure on $P$ by
$\mathcal M$ eliminates imaginaries. As a corollary, we obtain that every small
set $X$ definable in $\widetilde{\mathcal M}$ can be definably embedded into
some $P^l$, uniformly in parameters, settling a question from [10]. We verify
the tameness conditions in three examples: dense pairs of real closed fields,
expansions of $\mathcal M$ by a dense independent set, and expansions by a
dense divisible multiplicative group with the Mann property. Along the way, we
point out a gap in the proof of a relevant elimination of imaginaries result in
Wencel [17]. The above results are in contrast to recent literature, as it is
known in general that $\widetilde{\mathcal M}$ does not eliminate imaginaries,
and neither it nor the induced structure on $P$ admits definable Skolem
functions. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Enhanced ferromagnetic transition temperature induced by a microscopic structural rearrangement in the diluted magnetic semiconductor Ge$_{1-x}$Mn$_{x}$Te,
Abstract: The correlation between magnetic properties and microscopic structural
aspects in the diluted magnetic semiconductor Ge$_{1-x}$Mn$_{x}$Te is
investigated by x-ray diffraction and magnetization as a function of the Mn
concentration $x$. The occurrence of high ferromagnetic-transition temperatures
in the rhombohedrally distorted phase of slowly-cooled Ge$_{1-x}$Mn$_{x}$Te is
shown to be directly correlated with the formation and coexistence of
strongly-distorted Mn-poor and weakly-distorted Mn-rich regions. It is
demonstrated that the weakly-distorted phase fraction is responsible for the
occurrence of high-transition temperatures in Ge$_{1-x}$Mn$_{x}$Te. When the Mn
concentration becomes larger, the Mn-rich regions start to switch into the
undistorted cubic structure, and the transition temperature is suppressed
concurrently. By identifying suitable annealing conditions, we successfully
increased the transition temperature to above 200 K for Mn concentrations close
to the cubic phase. Structural data indicate that the weakly-distorted phase
fraction can be restored at the expense of the cubic regions upon the
enhancement of the transition temperature, clearly establishing the direct link
between high-transition temperatures and the weakly-distorted Mn-rich phase
fraction. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Courcelle's Theorem Made Dynamic,
Abstract: Dynamic complexity is concerned with updating the output of a problem when
the input is slightly changed. We study the dynamic complexity of model
checking a fixed monadic second-order formula over evolving subgraphs of a
fixed maximal graph having bounded tree-width; here the subgraph evolves by
losing or gaining edges (from the maximal graph). We show that this problem is
in DynFO (with LOGSPACE precomputation), via a reduction to a Dyck reachability
problem on an acyclic automaton. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: PorePy: An Open-Source Simulation Tool for Flow and Transport in Deformable Fractured Rocks,
Abstract: Fractures are ubiquitous in the subsurface and strongly affect flow and
deformation. The physical shape of the fractures, they are long and thin
objects, puts strong limitations on how the effect of this dynamics can be
incorporated into standard reservoir simulation tools. This paper reports the
development of an open-source software framework, termed PorePy, which is aimed
at simulation of flow and transport in three-dimensional fractured reservoirs,
as well as deformation of the reservoir due to shearing along fracture and
fault planes. Starting from a description of fractures as polygons embedded in
a 3D domain, PorePy provides semi-automatic gridding to construct a
discrete-fracture-matrix model, which forms the basis for subsequent
simulations. PorePy allows for flow and transport in all lower-dimensional
objects, including planes (2D) representing fractures, and lines (1D) and
points (0D), representing fracture intersections. Interaction between processes
in neighboring domains of different dimension is implemented as a sequence of
couplings of objects one dimension apart. This readily allows for handling of
complex fracture geometries compared to capabilities of existing software. In
addition to flow and transport, PorePy provides models for rock mechanics,
poro-elasticity and coupling with fracture deformation models. The software is
fully open, and can serve as a framework for transparency and reproducibility
of simulations. We describe the design principles of PorePy from a user
perspective, with focus on possibilities within gridding, covered physical
processes and available discretizations. The power of the framework is
illustrated with two sets of simulations; involving respectively coupled flow
and transport in a fractured porous medium, and low-pressure stimulation of a
geothermal reservoir. | [
1,
1,
0,
0,
0,
0
] | [
"Physics",
"Computer Science"
] |
Title: A Fast Integrated Planning and Control Framework for Autonomous Driving via Imitation Learning,
Abstract: For safe and efficient planning and control in autonomous driving, we need a
driving policy which can achieve desirable driving quality in long-term horizon
with guaranteed safety and feasibility. Optimization-based approaches, such as
Model Predictive Control (MPC), can provide such optimal policies, but their
computational complexity is generally unacceptable for real-time
implementation. To address this problem, we propose a fast integrated planning
and control framework that combines learning- and optimization-based approaches
in a two-layer hierarchical structure. The first layer, defined as the "policy
layer", is established by a neural network which learns the long-term optimal
driving policy generated by MPC. The second layer, called the "execution
layer", is a short-term optimization-based controller that tracks the reference
trajecotries given by the "policy layer" with guaranteed short-term safety and
feasibility. Moreover, with efficient and highly-representative features, a
small-size neural network is sufficient in the "policy layer" to handle many
complicated driving scenarios. This renders online imitation learning with
Dataset Aggregation (DAgger) so that the performance of the "policy layer" can
be improved rapidly and continuously online. Several exampled driving scenarios
are demonstrated to verify the effectiveness and efficiency of the proposed
framework. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: Predicted novel insulating electride compound between alkali metals lithium and sodium under high pressure,
Abstract: The application of high pressure can fundamentally modify the crystalline and
electronic structures of elements as well as their chemical reactivity, which
could lead to the formation of novel materials. Here, we explore the reactivity
of lithium with sodium under high pressure, using a swarm structure searching
techniques combined with first-principles calculations, which identify a
thermodynamically stable LiNa compound adopting an orthorhombic oP8 phase at
pressure above 355 GPa. The formation of LiNa may be a consequence of strong
concentration of electrons transfer from the lithium and the sodium atoms into
the interstitial sites, which also leads to opening a relatively wide band gap
for LiNa-op8. This is substantially different from the picture that share or
exchange electrons in common compounds and alloys. In addition, lattice-dynamic
calculations indicate that LiNa-op8 remains dynamically stable when pressure
decompresses down to 70 GPa. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: The transition matrix between the Specht and web bases is unipotent with additional vanishing entries,
Abstract: We compare two important bases of an irreducible representation of the
symmetric group: the web basis and the Specht basis. The web basis has its
roots in the Temperley-Lieb algebra and knot-theoretic considerations. The
Specht basis is a classic algebraic and combinatorial construction of symmetric
group representations which arises in this context through the geometry of
varieties called Springer fibers. We describe a graph that encapsulates
combinatorial relations between each of these bases, prove that there is a
unique way (up to scaling) to map the Specht basis into the web representation,
and use this to recover a result of Garsia-McLarnan that the transition matrix
between the Specht and web bases is upper-triangular with ones along the
diagonal. We then strengthen their result to prove vanishing of certain
additional entries unless a nesting condition on webs is satisfied. In fact we
conjecture that the entries of the transition matrix are nonnegative and are
nonzero precisely when certain directed paths exist in the web graph. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Robotic Wireless Sensor Networks,
Abstract: In this chapter, we present a literature survey of an emerging, cutting-edge,
and multi-disciplinary field of research at the intersection of Robotics and
Wireless Sensor Networks (WSN) which we refer to as Robotic Wireless Sensor
Networks (RWSN). We define a RWSN as an autonomous networked multi-robot system
that aims to achieve certain sensing goals while meeting and maintaining
certain communication performance requirements, through cooperative control,
learning and adaptation. While both of the component areas, i.e., Robotics and
WSN, are very well-known and well-explored, there exist a whole set of new
opportunities and research directions at the intersection of these two fields
which are relatively or even completely unexplored. One such example would be
the use of a set of robotic routers to set up a temporary communication path
between a sender and a receiver that uses the controlled mobility to the
advantage of packet routing. We find that there exist only a limited number of
articles to be directly categorized as RWSN related works whereas there exist a
range of articles in the robotics and the WSN literature that are also relevant
to this new field of research. To connect the dots, we first identify the core
problems and research trends related to RWSN such as connectivity,
localization, routing, and robust flow of information. Next, we classify the
existing research on RWSN as well as the relevant state-of-the-arts from
robotics and WSN community according to the problems and trends identified in
the first step. Lastly, we analyze what is missing in the existing literature,
and identify topics that require more research attention in the future. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: Using Mode Connectivity for Loss Landscape Analysis,
Abstract: Mode connectivity is a recently introduced frame- work that empirically
establishes the connected- ness of minima by finding a high accuracy curve
between two independently trained models. To investigate the limits of this
setup, we examine the efficacy of this technique in extreme cases where the
input models are trained or initialized differently. We find that the procedure
is resilient to such changes. Given this finding, we propose using the
framework for analyzing loss surfaces and training trajectories more generally,
and in this direction, study SGD with cosine annealing and restarts (SGDR). We
report that while SGDR moves over barriers in its trajectory, propositions
claiming that it converges to and escapes from multiple local minima are not
substantiated by our empirical results. | [
0,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Lifelong Generative Modeling,
Abstract: Lifelong learning is the problem of learning multiple consecutive tasks in a
sequential manner where knowledge gained from previous tasks is retained and
used for future learning. It is essential towards the development of
intelligent machines that can adapt to their surroundings. In this work we
focus on a lifelong learning approach to generative modeling where we
continuously incorporate newly observed distributions into our learnt model. We
do so through a student-teacher Variational Autoencoder architecture which
allows us to learn and preserve all the distributions seen so far without the
need to retain the past data nor the past models. Through the introduction of a
novel cross-model regularizer, inspired by a Bayesian update rule, the student
model leverages the information learnt by the teacher, which acts as a summary
of everything seen till now. The regularizer has the additional benefit of
reducing the effect of catastrophic interference that appears when we learn
over sequences of distributions. We demonstrate its efficacy in learning
sequentially observed distributions as well as its ability to learn a common
latent representation across a complex transfer learning scenario. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Non-convex Finite-Sum Optimization Via SCSG Methods,
Abstract: We develop a class of algorithms, as variants of the stochastically
controlled stochastic gradient (SCSG) methods (Lei and Jordan, 2016), for the
smooth non-convex finite-sum optimization problem. Assuming the smoothness of
each component, the complexity of SCSG to reach a stationary point with
$\mathbb{E} \|\nabla f(x)\|^{2}\le \epsilon$ is $O\left (\min\{\epsilon^{-5/3},
\epsilon^{-1}n^{2/3}\}\right)$, which strictly outperforms the stochastic
gradient descent. Moreover, SCSG is never worse than the state-of-the-art
methods based on variance reduction and it significantly outperforms them when
the target accuracy is low. A similar acceleration is also achieved when the
functions satisfy the Polyak-Lojasiewicz condition. Empirical experiments
demonstrate that SCSG outperforms stochastic gradient methods on training
multi-layers neural networks in terms of both training and validation loss. | [
1,
0,
1,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: Minority carrier diffusion lengths and mobilities in low-doped n-InGaAs for focal plane array applications,
Abstract: The hole diffusion length in n-InGaAs is extracted for two samples of
different doping concentrations using a set of long and thin diffused junction
diodes separated by various distances on the order of the diffusion length. The
methodology is described, including the ensuing analysis which yields diffusion
lengths between 70 - 85 um at room temperature for doping concentrations in the
range of 5 - 9 x 10^15 cm-3. The analysis also provides insight into the
minority carrier mobility which is a parameter not commonly reported in the
literature. Hole mobilities on the order of 500 - 750 cm2/Vs are reported for
the aforementioned doping range, which are comparable albeit longer than the
majority hole mobility for the same doping magnitude in p-InGaAs. A radiative
recombination coefficient of (0.5-0.2)x10^-10 cm^-3s^-1 is also extracted from
the ensuing analysis for an InGaAs thickness of 2.7 um. Preliminary evidence is
also given for both heavy and light hole diffusion. The dark current of
InP/InGaAs p-i-n photodetectors with 25 and 15 um pitches are then calibrated
to device simulations and correlated to the extracted diffusion lengths and
doping concentrations. An effective Shockley-Read-Hall lifetime of between
90-200 us provides the best fit to the dark current of these structures. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: A new Weber type integral equation related to the Weber-Titchmarsh problem,
Abstract: We derive solvability conditions and closed-form solution for the Weber type
integral equation, related to the familiar Weber-Orr integral transforms and
the old Weber-Titchmarsh problem (posed in Proc. Lond. Math. Soc. 22 (2)
(1924), pp.15, 16), recently solved by the author. Our method involves
properties of the inverse Mellin transform of integrable functions. The
Mellin-Parseval equality and some integrals, involving the Gauss hypergeometric
function are used. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Gate-Variants of Gated Recurrent Unit (GRU) Neural Networks,
Abstract: The paper evaluates three variants of the Gated Recurrent Unit (GRU) in
recurrent neural networks (RNN) by reducing parameters in the update and reset
gates. We evaluate the three variant GRU models on MNIST and IMDB datasets and
show that these GRU-RNN variant models perform as well as the original GRU RNN
model while reducing the computational expense. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science"
] |
Title: Some results on the annihilators and attached primes of local cohomology modules,
Abstract: Let $(R, \frak m)$ be a local ring and $M$ a finitely generated $R$-module.
It is shown that if $M$ is relative Cohen-Macaulay with respect to an ideal
$\frak a$ of $R$, then $\text{Ann}_R(H_{\mathfrak{a}}^{\text{cd}(\mathfrak{a},
M)}(M))=\text{Ann}_RM/L=\text{Ann}_RM$ and
$\text{Ass}_R(R/\text{Ann}_RM)\subseteq \{\mathfrak{p} \in \text{Ass}_R
M|\,{\rm cd}(\mathfrak{a}, R/\mathfrak{p})=\text{cd}(\mathfrak{a}, M)\},$ where
$L$ is the largest submodule of $M$ such that ${\rm cd}(\mathfrak{a}, L)< {\rm
cd}(\mathfrak{a}, M)$. We also show that if $H^{\dim M}_{\mathfrak{a}}(M)=0$,
then $\text{Att}_R(H^{\dim M-1}_{\mathfrak{a}}(M))= \{\mathfrak{p} \in
\text{Supp} (M)|\,{\rm cd}(\mathfrak{a}, R/\mathfrak{p})=\dim M-1\},$ and so
the attached primes of $H^{\dim M-1}_{\mathfrak{a}}(M)$ depends only on
$\text{Supp} (M)$. Finally, we prove that if $M$ is an arbitrary module (not
necessarily finitely generated) over a Noetherian ring $R$ with ${\rm
cd}(\mathfrak{a}, M)={\rm cd}(\mathfrak{a}, R/\text{Ann}_RM)$, then
$\text{Att}_R(H^{{\rm cd}(\mathfrak{a},
M)}_{\mathfrak{a}}(M))\subseteq\{\mathfrak{p} \in V(\text{Ann}_RM)|\,{\rm
cd}(\mathfrak{a}, R/\mathfrak{p})={\rm cd}(\mathfrak{a}, M)\}.$
As a consequence of this it is shown that if $\dim M=\dim R$, then
$\text{Att}_R(H^{\dim M}_{\mathfrak{a}}(M))\subseteq\{\mathfrak{p} \in
\text{Ass}_R M|\,{\rm cd}(\mathfrak{a}, R/\mathfrak{p})=\dim M\}.$ | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: A theoretical analysis of extending frequency-bin entanglement from photon-photon to atom-photon hybrid systems,
Abstract: Inspired by the recent developments in the research of atom-photon quantum
interface and energy-time entanglement between single photon pulses, we propose
to establish the concept of a special energy-time entanglement between a single
photon pulse and internal states of a single atom, which is analogous to the
frequency-bin entanglement between single photon pulses. We show that this type
of entanglement arises naturally in the interaction between frequency-bin
entangled single photon pulse pair and a single atom, via straightforward
atom-photon phase gate operations. We also discuss the properties of this type
of entanglement and show a preliminary example of its potential application in
quantum networking. Moreover, a quantum entanglement witness is constructed to
detect such entanglement from a reasonably large set of separable states. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Linearity of stability conditions,
Abstract: We study different concepts of stability for modules over a finite
dimensional algebra: linear stability, given by a "central charge", and
nonlinear stability given by the wall-crossing sequence of a "green path". Two
other concepts, finite Harder-Narasimhan stratification of the module category
and maximal forward hom-orthogonal sequences of Schurian modules, which are
always equivalent to each other, are shown to be equivalent to nonlinear
stability and to a maximal green sequence, defined using Fomin-Zelevinsky
quiver mutation, in the case the algebra is hereditary.
This is the first of a series of three papers whose purpose is to determine
all maximal green sequences of maximal length for quivers of affine type
$\tilde A$ and determine which are linear. The complete answer will be given in
the final paper [1]. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: BinPro: A Tool for Binary Source Code Provenance,
Abstract: Enforcing open source licenses such as the GNU General Public License (GPL),
analyzing a binary for possible vulnerabilities, and code maintenance are all
situations where it is useful to be able to determine the source code
provenance of a binary. While previous work has either focused on computing
binary-to-binary similarity or source-to-source similarity, BinPro is the first
work we are aware of to tackle the problem of source-to-binary similarity.
BinPro can match binaries with their source code even without knowing which
compiler was used to produce the binary, or what optimization level was used
with the compiler. To do this, BinPro utilizes machine learning to compute
optimal code features for determining binary-to-source similarity and a static
analysis pipeline to extract and compute similarity based on those features.
Our experiments show that on average BinPro computes a similarity of 81% for
matching binaries and source code of the same applications, and an average
similarity of 25% for binaries and source code of similar but different
applications. This shows that BinPro's similarity score is useful for
determining if a binary was derived from a particular source code. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: The placement of the head that maximizes predictability. An information theoretic approach,
Abstract: The minimization of the length of syntactic dependencies is a
well-established principle of word order and the basis of a mathematical theory
of word order. Here we complete that theory from the perspective of information
theory, adding a competing word order principle: the maximization of
predictability of a target element. These two principles are in conflict: to
maximize the predictability of the head, the head should appear last, which
maximizes the costs with respect to dependency length minimization. The
implications of such a broad theoretical framework to understand the
optimality, diversity and evolution of the six possible orderings of subject,
object and verb are reviewed. | [
1,
1,
0,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: Nonparametric regression using deep neural networks with ReLU activation function,
Abstract: Consider the multivariate nonparametric regression model. It is shown that
estimators based on sparsely connected deep neural networks with ReLU
activation function and properly chosen network architecture achieve the
minimax rates of convergence (up to log n-factors) under a general composition
assumption on the regression function. The framework includes many well-studied
structural constraints such as (generalized) additive models. While there is a
lot of flexibility in the network architecture, the tuning parameter is the
sparsity of the network. Specifically, we consider large networks with number
of potential network parameters exceeding the sample size. The analysis gives
some insights why multilayer feedforward neural networks perform well in
practice. Interestingly, the depth (number of layers) of the neural network
architectures plays an important role and our theory suggests that for
nonparametric regression scaling the network depth with the logarithm of the
sample size is natural. It is also shown that under the composition assumption
wavelet estimators can only achieve suboptimal rates. | [
1,
0,
1,
1,
0,
0
] | [
"Statistics",
"Computer Science",
"Mathematics"
] |
Title: BézierGAN: Automatic Generation of Smooth Curves from Interpretable Low-Dimensional Parameters,
Abstract: Many real-world objects are designed by smooth curves, especially in the
domain of aerospace and ship, where aerodynamic shapes (e.g., airfoils) and
hydrodynamic shapes (e.g., hulls) are designed. To facilitate the design
process of those objects, we propose a deep learning based generative model
that can synthesize smooth curves. The model maps a low-dimensional latent
representation to a sequence of discrete points sampled from a rational
Bézier curve. We demonstrate the performance of our method in completing both
synthetic and real-world generative tasks. Results show that our method can
generate diverse and realistic curves, while preserving consistent shape
variation in the latent space, which is favorable for latent space design
optimization or design space exploration. | [
0,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: Raman scattering study of tetragonal magnetic phase in Sr$_{1-x}$Na$_x$Fe$_2$As$_2$: structural symmetry and electronic gap,
Abstract: We use inelastic light scattering to study Sr$_{1-x}$Na$_x$Fe$_2$As$_2$
($x\approx0.34$), which exhibits a robust tetragonal magnetic phase that
restores the four-fold rotation symmetry inside the orthorhombic magnetic
phase. With cooling, we observe splitting and recombination of an $E_g$ phonon
peak upon entering the orthorhombic and tetragonal magnetic phases,
respectively, consistent with the reentrant phase behavior. Our electronic
Raman data reveal a pronounced feature that is clearly associated with the
tetragonal magnetic phase, suggesting the opening of an electronic gap. No
phonon back-folding behavior can be detected above the noise level, which
implies that any lattice translation symmetry breaking in the tetragonal
magnetic phase must be very weak. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Learning Mixture of Gaussians with Streaming Data,
Abstract: In this paper, we study the problem of learning a mixture of Gaussians with
streaming data: given a stream of $N$ points in $d$ dimensions generated by an
unknown mixture of $k$ spherical Gaussians, the goal is to estimate the model
parameters using a single pass over the data stream. We analyze a streaming
version of the popular Lloyd's heuristic and show that the algorithm estimates
all the unknown centers of the component Gaussians accurately if they are
sufficiently separated. Assuming each pair of centers are $C\sigma$ distant
with $C=\Omega((k\log k)^{1/4}\sigma)$ and where $\sigma^2$ is the maximum
variance of any Gaussian component, we show that asymptotically the algorithm
estimates the centers optimally (up to constants); our center separation
requirement matches the best known result for spherical Gaussians
\citep{vempalawang}. For finite samples, we show that a bias term based on the
initial estimate decreases at $O(1/{\rm poly}(N))$ rate while variance
decreases at nearly optimal rate of $\sigma^2 d/N$.
Our analysis requires seeding the algorithm with a good initial estimate of
the true cluster centers for which we provide an online PCA based clustering
algorithm. Indeed, the asymptotic per-step time complexity of our algorithm is
the optimal $d\cdot k$ while space complexity of our algorithm is $O(dk\log
k)$.
In addition to the bias and variance terms which tend to $0$, the
hard-thresholding based updates of streaming Lloyd's algorithm is agnostic to
the data distribution and hence incurs an approximation error that cannot be
avoided. However, by using a streaming version of the classical
(soft-thresholding-based) EM method that exploits the Gaussian distribution
explicitly, we show that for a mixture of two Gaussians the true means can be
estimated consistently, with estimation error decreasing at nearly optimal
rate, and tending to $0$ for $N\rightarrow \infty$. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Green-Blue Stripe Pattern for Range Sensing from a Single Image,
Abstract: In this paper, we present a novel method for rapid high-resolution range
sensing using green-blue stripe pattern. We use green and blue for designing
high-frequency stripe projection pattern. For accurate and reliable range
recovery, we identify the stripe patterns by our color-stripe segmentation and
unwrapping algorithms. The experimental result for a naked human face shows the
effectiveness of our method. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: Electrostatic gyrokinetic simulation of global tokamak boundary plasma and the generation of nonlinear intermittent turbulence,
Abstract: Boundary plasma physics plays an important role in tokamak confinement, but
is difficult to simulate in a gyrokinetic code due to the scale-inseparable
nonlocal multi-physics in magnetic separatrix and open magnetic field geometry.
Neutral particles are also an important part of the boundary plasma physics. In
the present paper, noble electrostatic gyrokinetic techniques to simulate the
flux-driven, low-beta electrostatic boundary plasma is reported. Gyrokinetic
ions and drift-kinetic electrons are utilized without scale-separation between
the neoclassical and turbulence dynamics. It is found that the nonlinear
intermittent turbulence is a natural gyrokinetic phenomenon in the boundary
plasma in the vicinity of the magnetic separatrix surface and in the scrape-off
layer. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Enhancing Interpretability of Black-box Soft-margin SVM by Integrating Data-based Priors,
Abstract: The lack of interpretability often makes black-box models difficult to be
applied to many practical domains. For this reason, the current work, from the
black-box model input port, proposes to incorporate data-based prior
information into the black-box soft-margin SVM model to enhance its
interpretability. The concept and incorporation mechanism of data-based prior
information are successively developed, based on which the interpretable or
partly interpretable SVM optimization model is designed and then solved through
handily rewriting the optimization problem as a nonlinear quadratic programming
problem. An algorithm for mining data-based linear prior information from data
set is also proposed, which generates a linear expression with respect to two
appropriate inputs identified from all inputs of system. At last, the proposed
interpretability enhancement strategy is applied to eight benchmark examples
for effectiveness exhibition. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: A Kuroda-style j-translation,
Abstract: In topos theory it is well-known that any nucleus j gives rise to a
translation of intuitionistic logic into itself in a way which generalises the
Goedel-Gentzen negative translation. Here we show that there exists a similar
j-translation which is more in the spirit of Kuroda's negative translation. The
key is to apply the nucleus not only to the entire formula and universally
quantified subformulas, but to conclusions of implications as well. The
development is entirely syntactic and no knowledge of topos theory is required
to read this small note. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Electrical 2π phase control of infrared light in a 350nm footprint using graphene plasmons,
Abstract: Modulating the amplitude and phase of light is at the heart of many
applications such as wavefront shaping, transformation optics, phased arrays,
modulators and sensors. Performing this task with high efficiency and small
footprint is a formidable challenge. Metasurfaces and plasmonics are promising
, but metals exhibit weak electro-optic effects. Two-dimensional materials,
such as graphene, have shown great performance as modulators with small drive
voltages. Here we show a graphene plasmonic phase modulator which is capable of
tuning the phase between 0 and 2{\pi} in situ. With a footprint of 350nm it is
more than 30 times smaller than the 10.6$\mu$m free space wavelength. The
modulation is achieved by spatially controlling the plasmon phase velocity in a
device where the spatial carrier density profile is tunable. We provide a
scattering theory for plasmons propagating through spatial density profiles.
This work constitutes a first step towards two-dimensional transformation
optics for ultra-compact modulators and biosensing. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Viscous dissipation of surface waves and its relevance to analogue gravity experiments,
Abstract: We consider dissipation of surface waves on fluids, with a view to its
effects on analogue gravity experiments. We begin by reviewing some general
properties of wave dissipation, before restricting our attention to surface
waves and the dissipative role played by viscosity there. Finally, with
particular focus on water, we consider several experimental setups inspired by
analogue gravity: the analogue Hawking effect, the black hole laser, the
analogue wormhole, and double bouncing at the wormhole entrance. Dissipative
effects are considered in each, and we give estimates for their optimized
experimental parameters. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Neural Optimizer Search with Reinforcement Learning,
Abstract: We present an approach to automate the process of discovering optimization
methods, with a focus on deep learning architectures. We train a Recurrent
Neural Network controller to generate a string in a domain specific language
that describes a mathematical update equation based on a list of primitive
functions, such as the gradient, running average of the gradient, etc. The
controller is trained with Reinforcement Learning to maximize the performance
of a model after a few epochs. On CIFAR-10, our method discovers several update
rules that are better than many commonly used optimizers, such as Adam,
RMSProp, or SGD with and without Momentum on a ConvNet model. We introduce two
new optimizers, named PowerSign and AddSign, which we show transfer well and
improve training on a variety of different tasks and architectures, including
ImageNet classification and Google's neural machine translation system. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Hybrid Sterility Can Only be Primary When Acting as a Reproductive Barrier for Sympatric Speciation,,
Abstract: Parental gametes unite to form a zygote that develops into an adult with
gonads that, in turn, produce gametes. Interruption of this germinal cycle by
prezygotic or postzygotic reproductive barriers can result in two independent
cycles, each with the potential to evolve into a new species. When the
speciation process is complete, members of each species are fully
reproductively isolated from those of the other. During speciation a primary
barrier may be supported and eventually superceded by a later appearing
secondary barrier. For those holding certain cases of prezygotic isolation to
be primary (e.g. elephant cannot copulate with mouse), the onus is to show that
they had not been preceded over evolutionary time by periods of postzygotic
hybrid inviability (genically determined) or sterility (genically or
chromosomally determined). Likewise, the onus is upon those holding cases of
hybrid inviability to be primary (e.g. Dobzhansky-Muller epistatic
incompatibilities), to show that they had not been preceded by periods, however
brief, of hybrid sterility. The latter, when acting as a sympatric barrier
causing reproductive isolation, can only be primary. In many cases, hybrid
sterility may result from incompatibilities between parental chromosomes that
attempt to pair during meiosis in the gonad of their offspring
(Winge-Crowther-Bateson incompatibilities). While WCB incompatibilities have
long been observed on a microscopic scale, there is growing evidence for a role
of dispersed finer DNA sequence differences. | [
0,
0,
0,
0,
1,
0
] | [
"Quantitative Biology"
] |
Title: Point-Cloud-Based Aerial Fragmentation Analysis for Application in the Minerals Industry,
Abstract: This work investigates the application of Unmanned Aerial Vehicle (UAV)
technology for measurement of rock fragmentation without placement of scale
objects in the scene to determine image scale. Commonly practiced image-based
rock fragmentation analysis requires a technician to walk to a rock pile, place
a scale object of known size in the area of interest, and capture individual 2D
images. Our previous work has used UAV technology for the first time to acquire
real-time rock fragmentation data and has shown comparable quality of results;
however, it still required the (potentially dangerous) placement of scale
objects, and continued to make the assumption that the rock pile surface is
planar and that the scale objects lie on the surface plane. This work improves
our UAV-based approach to enable rock fragmentation measurement without
placement of scale objects and without the assumption of planarity. This is
achieved by first generating a point cloud of the rock pile from 2D images,
taking into account intrinsic and extrinsic camera parameters, and then taking
2D images for fragmentation analysis. This work represents an important step
towards automating post-blast rock fragmentation analysis. In experiments, a
rock pile with known size distribution was photographed by the UAV with and
without using scale objects. For fragmentation analysis without scale objects,
a point cloud of the rock pile was generated and used to compute image scale.
Comparison of the rock size distributions show that this point-cloud-based
method enables producing measurements with better or comparable accuracy
(within 10% of the ground truth) to the manual method with scale objects. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Physics"
] |
Title: Where Classification Fails, Interpretation Rises,
Abstract: An intriguing property of deep neural networks is their inherent
vulnerability to adversarial inputs, which significantly hinders their
application in security-critical domains. Most existing detection methods
attempt to use carefully engineered patterns to distinguish adversarial inputs
from their genuine counterparts, which however can often be circumvented by
adaptive adversaries. In this work, we take a completely different route by
leveraging the definition of adversarial inputs: while deceiving for deep
neural networks, they are barely discernible for human visions. Building upon
recent advances in interpretable models, we construct a new detection framework
that contrasts an input's interpretation against its classification. We
validate the efficacy of this framework through extensive experiments using
benchmark datasets and attacks. We believe that this work opens a new direction
for designing adversarial input detection methods. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science"
] |
Title: Prior matters: simple and general methods for evaluating and improving topic quality in topic modeling,
Abstract: Latent Dirichlet Allocation (LDA) models trained without stopword removal
often produce topics with high posterior probabilities on uninformative words,
obscuring the underlying corpus content. Even when canonical stopwords are
manually removed, uninformative words common in that corpus will still dominate
the most probable words in a topic. In this work, we first show how the
standard topic quality measures of coherence and pointwise mutual information
act counter-intuitively in the presence of common but irrelevant words, making
it difficult to even quantitatively identify situations in which topics may be
dominated by stopwords. We propose an additional topic quality metric that
targets the stopword problem, and show that it, unlike the standard measures,
correctly correlates with human judgements of quality. We also propose a
simple-to-implement strategy for generating topics that are evaluated to be of
much higher quality by both human assessment and our new metric. This approach,
a collection of informative priors easily introduced into most LDA-style
inference methods, automatically promotes terms with domain relevance and
demotes domain-specific stop words. We demonstrate this approach's
effectiveness in three very different domains: Department of Labor accident
reports, online health forum posts, and NIPS abstracts. Overall we find that
current practices thought to solve this problem do not do so adequately, and
that our proposal offers a substantial improvement for those interested in
interpreting their topics as objects in their own right. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Density of Analytic Polynomials in Abstract Hardy Spaces,
Abstract: Let $X$ be a separable Banach function space on the unit circle $\mathbb{T}$
and $H[X]$ be the abstract Hardy space built upon $X$. We show that the set of
analytic polynomials is dense in $H[X]$ if the Hardy-Littlewood maximal
operator is bounded on the associate space $X'$. This result is specified to
the case of variable Lebesgue spaces. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: A spectral approach to the linking number in the 3-torus,
Abstract: Given a closed Riemannian manifold and a pair of multi-curves in it, we give
a formula relating the linking number of the later to the spectral theory of
the Laplace operator acting on differential one forms. As an application, we
compute the linking number of any two multi-geodesics of the flat torus of
dimension 3, generalising a result of P. Dehornoy. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics",
"Physics"
] |
Title: Volkov-Pankratov states in topological heterojunctions,
Abstract: We show that a smooth interface between two insulators of opposite
topological Z2 indices possesses multiple surface states, both massless and
massive. While the massless surface state is non-degenerate, chiral and
insensitive to the interface potential, the massive surface states only appear
for a sufficiently smooth heterojunction. The surface states are particle-hole
symmetric and a voltage drop reveals their intrinsic relativistic nature,
similarly to Landau bands of Dirac electrons in a magnetic field. We discuss
the relevance of the massive Dirac surface states in recent ARPES and transport
experiments. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Sampling-based Estimation of In-degree Distribution with Applications to Directed Complex Networks,
Abstract: The focus of this work is on estimation of the in-degree distribution in
directed networks from sampling network nodes or edges. A number of sampling
schemes are considered, including random sampling with and without replacement,
and several approaches based on random walks with possible jumps. When sampling
nodes, it is assumed that only the out-edges of that node are visible, that is,
the in-degree of that node is not observed. The suggested estimation of the
in-degree distribution is based on two approaches. The inversion approach
exploits the relation between the original and sample in-degree distributions,
and can estimate the bulk of the in-degree distribution, but not the tail of
the distribution. The tail of the in-degree distribution is estimated through
an asymptotic approach, which itself has two versions: one assuming a power-law
tail and the other for a tail of general form. The two estimation approaches
are examined on synthetic and real networks, with good performance results,
especially striking for the asymptotic approach. | [
1,
0,
0,
0,
0,
0
] | [
"Statistics",
"Computer Science"
] |
Title: Data Reduction and Image Reconstruction Techniques for Non-Redundant Masking,
Abstract: The technique of non-redundant masking (NRM) transforms a conventional
telescope into an interferometric array. In practice, this provides a much
better constrained point spread function than a filled aperture and thus higher
resolution than traditional imaging methods. Here we describe an NRM data
reduction pipeline. We discuss strategies for NRM observations regarding
dithering patterns and calibrator selection. We describe relevant image
calibrations and use example Large Binocular Telescope datasets to show their
effects on the scatter in the Fourier measurements. We also describe the
various ways to calculate Fourier quantities, and discuss different calibration
strategies. We present the results of image reconstructions from simulated
observations where we adjust prior images, weighting schemes, and error bar
estimation. We compare two imaging algorithms and discuss implications for
reconstructing images from real observations. Finally, we explore how the
current state of the art compares to next generation Extremely Large
Telescopes. | [
0,
1,
0,
0,
0,
0
] | [
"Physics",
"Computer Science"
] |
Title: Noisy Networks for Exploration,
Abstract: We introduce NoisyNet, a deep reinforcement learning agent with parametric
noise added to its weights, and show that the induced stochasticity of the
agent's policy can be used to aid efficient exploration. The parameters of the
noise are learned with gradient descent along with the remaining network
weights. NoisyNet is straightforward to implement and adds little computational
overhead. We find that replacing the conventional exploration heuristics for
A3C, DQN and dueling agents (entropy reward and $\epsilon$-greedy respectively)
with NoisyNet yields substantially higher scores for a wide range of Atari
games, in some cases advancing the agent from sub to super-human performance. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Global stability of the Rate Control Protocol (RCP) and some implications for protocol design,
Abstract: The Rate Control Protocol (RCP) is a congestion control protocol that relies
on explicit feedback from routers. RCP estimates the flow rate using two forms
of feedback: rate mismatch and queue size. However, it remains an open design
question whether queue size feedback in RCP is useful, given the presence of
rate mismatch. The model we consider has RCP flows operating over a single
bottleneck, with heterogeneous time delays. We first derive a sufficient
condition for global stability, and then highlight how this condition favors
the design choice of having only rate mismatch in the protocol definition. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: Trading Bounds for Memory in Games with Counters,
Abstract: We study two-player games with counters, where the objective of the first
player is that the counter values remain bounded. We investigate the existence
of a trade-off between the size of the memory and the bound achieved on the
counters, which has been conjectured by Colcombet and Loeding.
We show that unfortunately this conjecture does not hold: there is no
trade-off between bounds and memory, even for finite arenas. On the positive
side, we prove the existence of a trade-off for the special case of thin tree
arenas. This allows to extend the theory of regular cost functions over thin
trees, and obtain as a corollary the decidability of cost monadic second-order
logic over thin trees. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: The Frobenius problem for four numerical semigroups,
Abstract: The greatest integer that does not belong to a numerical semigroup $S$ is
called the Frobenius number of $S$ and finding the Frobenius number is called
the Frobenius problem. In this paper, we introduce the Frobenius problem for
numerical semigroups generated by Thabit number base b and Thabit number of the
second kind base b which are motivated by the Frobenius problem for Thabit
numerical semigroups. Also, we introduce the Frobenius problem for numerical
semigroups generated by Cunningham number and Fermat number base $b$ | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Coherence and its Role in Excitation Energy Transfer in Fenna-Mathews-Olson Complex,
Abstract: We show that the coherence between different bacteriochlorophyll-a (BChla)
sites in the Fenna-Mathews-Olson complex is an essential ingredient for
excitation energy transfer between various sites. The coherence delocalizes the
excitation energy, which results in the redistribution of excitation among all
the BChla sites in the steady state. We further show that the system remains
partially coherent at the steady state. In our numerical simulation of the
non-Markovian density matrix equation, we consider both the inhomogeneity of
the protein environment and the effect of active vibronic modes. | [
0,
1,
0,
0,
0,
0
] | [
"Physics",
"Quantitative Biology"
] |
Title: Effective field theory for dissipative fluids (II): classical limit, dynamical KMS symmetry and entropy current,
Abstract: In this paper we further develop the fluctuating hydrodynamics proposed in
arXiv:1511.03646 in a number of ways. We first work out in detail the classical
limit of the hydrodynamical action, which exhibits many simplifications. In
particular, this enables a transparent formulation of the action in physical
spacetime in the presence of arbitrary external fields. It also helps to
clarify issues related to field redefinitions and frame choices. We then
propose that the action is invariant under a $Z_2$ symmetry to which we refer
as the dynamical KMS symmetry. The dynamical KMS symmetry is physically
equivalent to the previously proposed local KMS condition in the classical
limit, but is more convenient to implement and more general. It is applicable
to any states in local equilibrium rather than just thermal density matrix
perturbed by external background fields. Finally we elaborate the formulation
for a conformal fluid, which contains some new features, and work out the
explicit form of the entropy current to second order in derivatives for a
neutral conformal fluid. | [
0,
1,
1,
0,
0,
0
] | [
"Physics"
] |
Title: Quantum non demolition measurements: parameter estimation for mixtures of multinomials,
Abstract: In Quantum Non Demolition measurements, the sequence of observations is
distributed as a mixture of multinomial random variables. Parameters of the
dynamics are naturally encoded into this family of distributions. We show the
local asymptotic mixed normality of the underlying statistical model and the
consistency of the maximum likelihood estimator. Furthermore, we prove the
asymptotic optimality of this estimator as it saturates the usual Cramér Rao
bound. | [
0,
0,
1,
1,
0,
0
] | [
"Physics",
"Statistics"
] |
Title: Why We Need New Evaluation Metrics for NLG,
Abstract: The majority of NLG evaluation relies on automatic metrics, such as BLEU . In
this paper, we motivate the need for novel, system- and data-independent
automatic evaluation methods: We investigate a wide range of metrics, including
state-of-the-art word-based and novel grammar-based ones, and demonstrate that
they only weakly reflect human judgements of system outputs as generated by
data-driven, end-to-end NLG. We also show that metric performance is data- and
system-specific. Nevertheless, our results also suggest that automatic metrics
perform reliably at system-level and can support system development by finding
cases where a system performs poorly. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Quasars Probing Quasars IX. The Kinematics of the Circumgalactic Medium Surrounding z ~ 2 Quasars,
Abstract: We examine the kinematics of the gas in the environments of galaxies hosting
quasars at $z\sim2$. We employ 148 projected quasar pairs to study the
circumgalactic gas of the foreground quasars in absorption. The sample selects
foreground quasars with precise redshift measurements, using emission-lines
with precision $\lesssim300\,{\rm km\,s^{-1}}$ and average offsets from the
systemic redshift $\lesssim|100\,{\rm km\,s^{-1}}|$. We stack the background
quasar spectra at the foreground quasar's systemic redshift to study the mean
absorption in \ion{C}{2}, \ion{C}{4}, and \ion{Mg}{2}. We find that the mean
absorptions exhibit large velocity widths $\sigma_v\approx300\,{\rm
km\,s^{-1}}$. Further, the mean absorptions appear to be asymmetric about the
systemic redshifts. The mean absorption centroids exhibit small redshift
relative to the systemic $\delta v\approx+200\,{\rm km\,s^{-1}}$, with large
intrinsic scatter in the centroid velocities of the individual absorption
systems. We find the observed widths are consistent with gas in gravitational
motion and Hubble flow. However, while the observation of large widths alone
does not require galactic-scale outflows, the observed offsets suggest that the
gas is on average outflowing from the galaxy. The observed offsets also suggest
that the ionizing radiation from the foreground quasars is anisotropic and/or
intermittent. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Deep Learning with Permutation-invariant Operator for Multi-instance Histopathology Classification,
Abstract: The computer-aided analysis of medical scans is a longstanding goal in the
medical imaging field. Currently, deep learning has became a dominant
methodology for supporting pathologists and radiologist. Deep learning
algorithms have been successfully applied to digital pathology and radiology,
nevertheless, there are still practical issues that prevent these tools to be
widely used in practice. The main obstacles are low number of available cases
and large size of images (a.k.a. the small n, large p problem in machine
learning), and a very limited access to annotation at a pixel level that can
lead to severe overfitting and large computational requirements. We propose to
handle these issues by introducing a framework that processes a medical image
as a collection of small patches using a single, shared neural network. The
final diagnosis is provided by combining scores of individual patches using a
permutation-invariant operator (combination). In machine learning community
such approach is called a multi-instance learning (MIL). | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Quantitative Biology"
] |
Title: Reeb dynamics inspired by Katok's example in Finsler geometry,
Abstract: Inspired by Katok's examples of Finsler metrics with a small number of closed
geodesics, we present two results on Reeb flows with finitely many periodic
orbits. The first result is concerned with a contact-geometric description of
magnetic flows on the 2-sphere found recently by Benedetti. We give a simple
interpretation of that work in terms of a quaternionic symmetry. In the second
part, we use Hamiltonian circle actions on symplectic manifolds to produce
compact, connected contact manifolds in dimension at least five with
arbitrarily large numbers of periodic Reeb orbits. This contrasts sharply with
recent work by Cristofaro-Gardiner, Hutchings and Pomerleano on Reeb flows in
dimension three. With the help of Hamiltonian plugs and a surgery construction
due to Laudenbach we reprove a result of Cieliebak: one can produce Hamiltonian
flows in dimension at least five with any number of periodic orbits; in
dimension three, with any number greater than one. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics",
"Physics"
] |
Title: Perturbative Thermodynamic Geometry of Nonextensive Ideal Classical, Bose and Fermi Gases,
Abstract: We investigate perturbative thermodynamic geometry of nonextensive ideal
Classical, Bose and Fermi gases.We show that the intrinsic statistical
interaction of nonextensive Bose (Fermi) gas is attractive (repulsive) similar
to the extensive case but the value of thermodynamic curvature is changed by
nonextensive parameter. In contrary to the extensive ideal classical gas, the
nonextensive one may be divided to two different regimes. According to
deviation parameter of the system to the nonextensive case, one can find a
special value of fugacity, $z^{*}$, where the sign of thermodynamic curvature
is changed. Therefore, we argue that the nonextensive parameter induces an
attractive (repulsive) statistical interaction for $z<z^{*}$ ($z>z^{*}$) for an
ideal classical gas. Also, according to the singular point of thermodynamic
curvature, we consider the condensation of nonextensive Boson gas. | [
0,
1,
0,
0,
0,
0
] | [
"Physics",
"Mathematics"
] |
Title: Timelike surfaces in Minkowski space with a canonical null direction,
Abstract: Given a constant vector field $Z$ in Minkowski space, a timelike surface is
said to have a canonical null direction with respect to $Z$ if the projection
of $Z$ on the tangent space of the surface gives a lightlike vector field. In
this paper we describe these surfaces in the ruled case. For example when the
Minkowski space has three dimensions then a surface with a canonical null
direction is minimal and flat. On the other hand, we describe several
properties in the non ruled case and we partially describe these surfaces in
four-dimensional Minkowski space. We give different ways for building these
surfaces in four-dimensional Minkowski space and we finally use the Gauss map
for describe another properties of these surfaces. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics",
"Physics"
] |
Title: Fluid-Structure Interaction for the Classroom: Interpolation, Hearts, and Swimming!,
Abstract: While students may find spline interpolation easily digestible, based on
their familiarity with continuity of a function and its derivatives, some of
its inherent value may be missed when students only see it applied to standard
data interpolation exercises. In this paper, we offer alternatives where
students can qualitatively and quantitatively witness the resulting dynamical
differences when objects are driven through a fluid using different spline
interpolation methods. They say, seeing is believing; here we showcase the
differences between linear and cubic spline interpolation using examples from
fluid pumping and aquatic locomotion. Moreover, students can define their own
interpolation functions and visualize the dynamics unfold. To solve the
fluid-structure interaction system, the open source software IB2d is used. In
that vein, all simulation codes, analysis scripts, and movies are provided for
streamlined use. | [
0,
0,
0,
0,
1,
0
] | [
"Physics",
"Mathematics"
] |
Title: A cellular algebra with specific decomposition of the unity,
Abstract: Let $ \mathbb{A}$ be a cellular algebra over a field $\mathbb{F}$ with a
decomposition of the identity $ 1_{\mathbb{A}} $ into orthogonal idempotents $
e_i$, $i \in I$ (for some finite set $I$) satisfying some properties. We
describe the entire Loewy structure of cell modules of the algebra $ \mathbb{A}
$ by using the representation theory of the algebra $ e_i \mathbb{A} e_i $ for
each $ i $. Moreover, we also study the block theory of $\mathbb{A}$ by using
this decomposition. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Distance-based classifier by data transformation for high-dimension, strongly spiked eigenvalue models,
Abstract: We consider classifiers for high-dimensional data under the strongly spiked
eigenvalue (SSE) model. We first show that high-dimensional data often have the
SSE model. We consider a distance-based classifier using eigenstructures for
the SSE model. We apply the noise reduction methodology to estimation of the
eigenvalues and eigenvectors in the SSE model. We create a new distance-based
classifier by transforming data from the SSE model to the non-SSE model. We
give simulation studies and discuss the performance of the new classifier.
Finally, we demonstrate the new classifier by using microarray data sets. | [
0,
0,
0,
1,
0,
0
] | [
"Statistics",
"Computer Science"
] |
Title: Group Field theory and Tensor Networks: towards a Ryu-Takayanagi formula in full quantum gravity,
Abstract: We establish a dictionary between group field theory (thus, spin networks and
random tensors) states and generalized random tensor networks. Then, we use
this dictionary to compute the Rényi entropy of such states and recover the
Ryu-Takayanagi formula, in two different cases corresponding to two different
truncations/approximations, suggested by the established correspondence. | [
0,
1,
0,
0,
0,
0
] | [
"Physics",
"Mathematics"
] |
Title: Behavior of Accelerated Gradient Methods Near Critical Points of Nonconvex Functions,
Abstract: We examine the behavior of accelerated gradient methods in smooth nonconvex
unconstrained optimization, focusing in particular on their behavior near
strict saddle points. Accelerated methods are iterative methods that typically
step along a direction that is a linear combination of the previous step and
the gradient of the function evaluated at a point at or near the current
iterate. (The previous step encodes gradient information from earlier stages in
the iterative process.) We show by means of the stable manifold theorem that
the heavy-ball method method is unlikely to converge to strict saddle points,
which are points at which the gradient of the objective is zero but the Hessian
has at least one negative eigenvalue. We then examine the behavior of the
heavy-ball method and other accelerated gradient methods in the vicinity of a
strict saddle point of a nonconvex quadratic function, showing that both
methods can diverge from this point more rapidly than the steepest-descent
method. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics",
"Computer Science"
] |
Title: Threat analysis of IoT networks Using Artificial Neural Network Intrusion Detection System,
Abstract: The Internet of things (IoT) is still in its infancy and has attracted much
interest in many industrial sectors including medical fields, logistics
tracking, smart cities and automobiles. However as a paradigm, it is
susceptible to a range of significant intrusion threats. This paper presents a
threat analysis of the IoT and uses an Artificial Neural Network (ANN) to
combat these threats. A multi-level perceptron, a type of supervised ANN, is
trained using internet packet traces, then is assessed on its ability to thwart
Distributed Denial of Service (DDoS/DoS) attacks. This paper focuses on the
classification of normal and threat patterns on an IoT Network. The ANN
procedure is validated against a simulated IoT network. The experimental
results demonstrate 99.4% accuracy and can successfully detect various DDoS/DoS
attacks. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: An experimental study of Bitcoin fluctuation using machine learning methods,
Abstract: In this paper, we study the ability to make the short-term prediction of the
exchange price fluctuations towards the United States dollar for the Bitcoin
market. We use the data of realized volatility collected from one of the
largest Bitcoin digital trading offices in 2016 and 2017 as well as order
information. Experiments are performed to evaluate a variety of statistical and
machine learning approaches. | [
0,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics",
"Quantitative Finance"
] |
Title: Radiation-driven turbulent accretion onto massive black holes,
Abstract: Accretion of gas and interaction of matter and radiation are at the heart of
many questions pertaining to black hole (BH) growth and coevolution of massive
BHs and their host galaxies. To answer them it is critical to quantify how the
ionizing radiation that emanates from the innermost regions of the BH accretion
flow couples to the surrounding medium and how it regulates the BH fueling. In
this work we use high resolution 3-dimensional (3D) radiation-hydrodynamic
simulations with the code Enzo, equipped with adaptive ray tracing module
Moray, to investigate radiation-regulated BH accretion of cold gas. Our
simulations reproduce findings from an earlier generation of 1D/2D simulations:
the accretion powered UV and X-ray radiation forms a highly ionized bubble,
which leads to suppression of BH accretion rate characterized by quasi-periodic
outbursts. A new feature revealed by the 3D simulations is the highly turbulent
nature of the gas flow in vicinity of the ionization front. During quiescent
periods between accretion outbursts, the ionized bubble shrinks in size and the
gas density that precedes the ionization front increases. Consequently, the 3D
simulations show oscillations in the accretion rate of only ~2-3 orders of
magnitude, significantly smaller than 1D/2D models. We calculate the energy
budget of the gas flow and find that turbulence is the main contributor to the
kinetic energy of the gas but corresponds to less than 10% of its thermal
energy and thus does not contribute significantly to the pressure support of
the gas. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Bayesian Inference of Log Determinants,
Abstract: The log-determinant of a kernel matrix appears in a variety of machine
learning problems, ranging from determinantal point processes and generalized
Markov random fields, through to the training of Gaussian processes. Exact
calculation of this term is often intractable when the size of the kernel
matrix exceeds a few thousand. In the spirit of probabilistic numerics, we
reinterpret the problem of computing the log-determinant as a Bayesian
inference problem. In particular, we combine prior knowledge in the form of
bounds from matrix theory and evidence derived from stochastic trace estimation
to obtain probabilistic estimates for the log-determinant and its associated
uncertainty within a given computational budget. Beyond its novelty and
theoretic appeal, the performance of our proposal is competitive with
state-of-the-art approaches to approximating the log-determinant, while also
quantifying the uncertainty due to budget-constrained evidence. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics",
"Mathematics"
] |
Title: Amplitude death and resurgence of oscillation in network of mobile oscillators,
Abstract: The phenomenon of amplitude death has been explored using a variety of
different coupling strategies in the last two decades. In most of the work, the
basic coupling arrangement is considered to be static over time, although many
realistic systems exhibit significant changes in the interaction pattern as
time varies. In this article, we study the emergence of amplitude death in a
dynamical network composed of time-varying interaction amidst a collection of
random walkers in a finite region of three dimensional space. We consider an
oscillator for each walker and demonstrate that depending upon the network
parameters and hence the interaction between them, global oscillation in the
network gets suppressed. In this framework, vision range of each oscillator
decides the number of oscillators with which it interacts. In addition, with
the use of an appropriate feedback parameter in the coupling strategy, we
articulate how the suppressed oscillation can be resurrected in the systems'
parameter space. The phenomenon of amplitude death and the resurgence of
oscillation is investigated taking limit cycle and chaotic oscillators for
broad ranges of parameters, like interaction strength k between the entities,
vision range r and the speed of movement v. | [
0,
1,
0,
0,
0,
0
] | [
"Physics",
"Mathematics"
] |
Title: Degenerate and chiral states in the extended Heisenberg model in the kagome lattice,
Abstract: We present a study of the low temperature phases of the antiferromagnetic
extended classical Heisenberg model in the kagome lattice, up to third nearest
neighbors. First, we focus on the degenerate lines in the boundaries of the
well-known staggered chiral phases. These boundaries have either semi-extensive
or extensive degeneracy, and we discuss the partial selection of states by
thermal fluctuations. Then, we study the model under an external magnetic field
on these lines and in the staggered chiral phases. We pay particular attention
to the highly frustrated point, where the three exchange couplings are equal.
We show that this point can me mapped to a model with spin liquid behavior and
non-zero chirality. Finally, we explore the effect of Dzyaloshinskii-Moriya
(DM) interactions in two ways: an homogeneous and a staggered DM interaction.
In both cases, there is a rich low temperature phase diagram, with different
spontaneously broken symmetries and non trivial chiral phases. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation,
Abstract: Generating novel graph structures that optimize given objectives while
obeying some given underlying rules is fundamental for chemistry, biology and
social science research. This is especially important in the task of molecular
graph generation, whose goal is to discover novel molecules with desired
properties such as drug-likeness and synthetic accessibility, while obeying
physical laws such as chemical valency. However, designing models to find
molecules that optimize desired properties while incorporating highly complex
and non-differentiable rules remains to be a challenging task. Here we propose
Graph Convolutional Policy Network (GCPN), a general graph convolutional
network based model for goal-directed graph generation through reinforcement
learning. The model is trained to optimize domain-specific rewards and
adversarial loss through policy gradient, and acts in an environment that
incorporates domain-specific rules. Experimental results show that GCPN can
achieve 61% improvement on chemical property optimization over state-of-the-art
baselines while resembling known molecules, and achieve 184% improvement on the
constrained property optimization task. | [
0,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Quantitative Biology"
] |
Title: Impact of the positive ion current on large size neutrino detectors and delayed photon emission,
Abstract: Given their small mobility coefficient in liquid argon with respect to the
electrons, the ions spend a considerably longer time in the active volume. We
studied the effects of the positive ion current in a liquid argon time
projection chamber, in the context of massive argon experiments for neutrino
physics. The constant recombination between free ions and electrons produces a
quenching of the charge signal and a constant emission of photons, uncorrelated
in time and space to the physical interactions. The predictions evidence some
potential concerns for multi-ton argon detectors, particularly when operated on
surface | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: On Multilevel Coding Schemes Based on Non-Binary LDPC Codes,
Abstract: We address the problem of constructing of coding schemes for the channels
with high-order modulations. It is known, that non-binary LDPC codes are
especially good for such channels and significantly outperform their binary
counterparts. Unfortunately, their decoding complexity is still large. In order
to reduce the decoding complexity we consider multilevel coding schemes based
on non-binary LDPC codes (NB-LDPC-MLC schemes) over smaller fields. The use of
such schemes gives us a reasonable gain in complexity. At the same time the
performance of NB-LDPC-MLC schemes is practically the same as the performance
of LDPC codes over the field matching the modulation order. In particular by
means of simulations we showed that the performance of NB-LDPC-MLC schemes over
GF(16) is the same as the performance of non-binary LDPC codes over GF(64) and
GF(256) in AWGN channel with QAM64 and QAM256 accordingly. We also perform a
comparison with binary LDPC codes. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Mathematics"
] |
Title: High-precision measurements and theoretical calculations of indium excited-state polarizabilities,
Abstract: We report measurements of the $^{115}$In $7p_{1/2}$ and $7p_{3/2}$ scalar and
tensor polarizabilities using two-step diode laser spectroscopy in an atomic
beam. The scalar polarizabilities are one to two orders of magnitude larger
than for lower lying indium states due to the close proximity of the $7p$ and
$6d$ states. For the scalar polarizabilities, we find values (in atomic units)
of $1.811(4) \times 10^5$ $a_0^3$ and $2.876(6) \times 10^5$ $a_0^3$ for the
$7p_{1/2}$ and $7p_{3/2}$ states respectively. We estimate the smaller tensor
polarizability component of the $7p_{3/2}$ state to be $-1.43(18) \times 10^4$
$a_0^3$. These measurements represent the first high-precision benchmarks of
transition properties of such high excited states of trivalent atomic systems.
We also present new ab initio calculations of these quantities and other In
polarizabilities using two high-precision relativistic methods to make a global
comparison of the accuracies of the two approaches. The precision of the
experiment is sufficient to differentiate between the two theoretical methods
as well as to allow precise determination of the indium $7p-6d$ matrix
elements. The results obtained in this work are applicable to other heavier and
more complicated systems, and provide much needed guidance for the development
of even more precise theoretical approaches. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Schur $Q$-functions and the Capelli eigenvalue problem for the Lie superalgebra $\mathfrak q(n)$,
Abstract: Let $\mathfrak l:= \mathfrak q(n)\times\mathfrak q(n)$, where $\mathfrak
q(n)$ denotes the queer Lie superalgebra. The associative superalgebra $V$ of
type $Q(n)$ has a left and right action of $\mathfrak q(n)$, and hence is
equipped with a canonical $\mathfrak l$-module structure. We consider a
distinguished basis $\{D_\lambda\}$ of the algebra of $\mathfrak l$-invariant
super-polynomial differential operators on $V$, which is indexed by strict
partitions of length at most $n$. We show that the spectrum of the operator
$D_\lambda$, when it acts on the algebra $\mathscr P(V)$ of super-polynomials
on $V$, is given by the factorial Schur $Q$-function of Okounkov and Ivanov.
This constitutes a refinement and a new proof of a result of Nazarov, who
computed the top-degree homogeneous part of the Harish-Chandra image of
$D_\lambda$. As a further application, we show that the radial projections of
the spherical super-polynomials corresponding to the diagonal symmetric pair
$(\mathfrak l,\mathfrak m)$, where $\mathfrak m:=\mathfrak q(n)$, of
irreducible $\mathfrak l$-submodules of $\mathscr P(V)$ are the classical Schur
$Q$-functions. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Dynamics of cosmological perturbations in modified Brans-Dicke cosmology with matter-scalar field interaction,
Abstract: In this work we focus on a novel completion of the well-known Brans-Dicke
theory that introduces an interaction between the dark energy and dark matter
sectors, known as complete Brans-Dicke (CBD) theory. We obtain viable
cosmological accelerating solutions that fit Supernovae observations with great
precision without any scalar potential $V(\phi)$. We use these solutions to
explore the impact of the CBD theory on the large scale structure by studying
the dynamics of its linear perturbations. We observe a growing behavior of the
lensing potential $\Phi_{+}$ at late-times, while the growth rate is actually
suppressed relatively to $\Lambda$CDM, which allows the CBD theory to provide a
competitive fit to current RSD measurements of $f\sigma_{8}$. However, we also
observe that the theory exhibits a pathological change of sign in the effective
gravitational constant concerning the perturbations on sub-horizon scales that
could pose a challenge to its validity. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Men Are from Mars, Women Are from Venus: Evaluation and Modelling of Verbal Associations,
Abstract: We present a quantitative analysis of human word association pairs and study
the types of relations presented in the associations. We put our main focus on
the correlation between response types and respondent characteristics such as
occupation and gender by contrasting syntagmatic and paradigmatic associations.
Finally, we propose a personalised distributed word association model and show
the importance of incorporating demographic factors into the models commonly
used in natural language processing. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Factorisation of the product of Dirichlet series of completely multiplicative functions,
Abstract: In the first chapter, we will present a computation of the square value of
the module of L functions associated to a Dirichlet character. This computation
suggests to ask if a certain ring of arithmetic multiplicative functions exists
and if it is unique. This search has led to the construction of that ring in
chapter two. Finally, in the third chapter, we will present some propositions
associated with this ring. The result below is one of the main results of this
work :
For F and G two completely multiplicative functions, $ s $ a complex number
such as the dirichlet series $ D(F,s) $ and $ D(G,s) $ converge :
$ \forall F,G \in \mathbb{M}_{c} : D(F,s) \times D(G,s) = D(F \times G,2s)
\times D(F \square G,s) $
where the operation $ \square $ is defined in chapter two as the sum of the
previously mentioned ring. Here are some similar versions, with $ s = x+iy $ :
$ \forall F, G \in \mathbb{M}_{c} : ~ D(F,s) \times D(G,\overline{s}) = D(F
\times G,2x) \times D(\frac{F}{\text{Id}_{e}^{iy}} \square
\frac{G}{\text{Id}_{e}^{-iy}}, x) $
$ \forall F, G \in \mathbb{M}_{c} : ~ |D(F,s)|^{2} = D(|F|^{2},2x) \times
D(\frac{F}{\text{Id}_{e}^{iy}} \square \overline{\frac{F}{\text{Id}_{e}^{iy}}},
x) $ | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: LAP: a Linearize and Project Method for Solving Inverse Problems with Coupled Variables,
Abstract: Many inverse problems involve two or more sets of variables that represent
different physical quantities but are tightly coupled with each other. For
example, image super-resolution requires joint estimation of the image and
motion parameters from noisy measurements. Exploiting this structure is key for
efficiently solving these large-scale optimization problems, which are often
ill-conditioned.
In this paper, we present a new method called Linearize And Project (LAP)
that offers a flexible framework for solving inverse problems with coupled
variables. LAP is most promising for cases when the subproblem corresponding to
one of the variables is considerably easier to solve than the other. LAP is
based on a Gauss-Newton method, and thus after linearizing the residual, it
eliminates one block of variables through projection. Due to the linearization,
this block can be chosen freely. Further, LAP supports direct, iterative, and
hybrid regularization as well as constraints. Therefore LAP is attractive,
e.g., for ill-posed imaging problems. These traits differentiate LAP from
common alternatives for this type of problem such as variable projection
(VarPro) and block coordinate descent (BCD). Our numerical experiments compare
the performance of LAP to BCD and VarPro using three coupled problems whose
forward operators are linear with respect to one block and nonlinear for the
other set of variables. | [
1,
0,
1,
0,
0,
0
] | [
"Mathematics",
"Computer Science"
] |
Title: Solitons and breathers for nonisospectral mKdV equation with Darboux transformation,
Abstract: Under investigation in this paper is the nonisospectral and variable
coefficients modified Kortweg-de Vries (vc-mKdV) equation, which manifests in
diverse areas of physics such as fluid dynamics, ion acoustic solitons and
plasma mechanics. With the degrees of restriction reduced, a simplified
constraint is introduced, under which the vc-mKdV equation is an integrable
system and the spectral flow is time-varying. The Darboux transformation for
such equation is constructed, which gives rise to the generation of variable
kinds of solutions including the double-breather coherent structure, periodical
soliton-breather and localized solitons and breathers. In addition, the effect
of variable coefficients and initial phases is discussed in terms of the
soliton amplitude, polarity, velocity and width, which might provide feasible
soliton management with certain conditions taken into account. | [
0,
1,
0,
0,
0,
0
] | [
"Physics",
"Mathematics"
] |
Title: Sequential detection of low-rank changes using extreme eigenvalues,
Abstract: We study the problem of detecting an abrupt change to the signal covariance
matrix. In particular, the covariance changes from a "white" identity matrix to
an unknown spiked or low-rank matrix. Two sequential change-point detection
procedures are presented, based on the largest and the smallest eigenvalues of
the sample covariance matrix. To control false-alarm-rate, we present an
accurate theoretical approximation to the average-run-length (ARL) and expected
detection delay (EDD) of the detection, leveraging the extreme eigenvalue
distributions from random matrix theory and by capturing a non-negligible
temporal correlation in the sequence of scan statistics due to the sliding
window approach. Real data examples demonstrate the good performance of our
method for detecting behavior change of a swarm. | [
0,
0,
1,
1,
0,
0
] | [
"Statistics",
"Mathematics"
] |
Title: Embodied Artificial Intelligence through Distributed Adaptive Control: An Integrated Framework,
Abstract: In this paper, we argue that the future of Artificial Intelligence research
resides in two keywords: integration and embodiment. We support this claim by
analyzing the recent advances of the field. Regarding integration, we note that
the most impactful recent contributions have been made possible through the
integration of recent Machine Learning methods (based in particular on Deep
Learning and Recurrent Neural Networks) with more traditional ones (e.g.
Monte-Carlo tree search, goal babbling exploration or addressable memory
systems). Regarding embodiment, we note that the traditional benchmark tasks
(e.g. visual classification or board games) are becoming obsolete as
state-of-the-art learning algorithms approach or even surpass human performance
in most of them, having recently encouraged the development of first-person 3D
game platforms embedding realistic physics. Building upon this analysis, we
first propose an embodied cognitive architecture integrating heterogenous
sub-fields of Artificial Intelligence into a unified framework. We demonstrate
the utility of our approach by showing how major contributions of the field can
be expressed within the proposed framework. We then claim that benchmarking
environments need to reproduce ecologically-valid conditions for bootstrapping
the acquisition of increasingly complex cognitive skills through the concept of
a cognitive arms race between embodied agents. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science"
] |
Title: Canonical models of arithmetic $(1; \infty)$ curves,
Abstract: In 1983 Takeuchi showed that up to conjugation there are exactly 4 arithmetic
subgroups of $\textrm{PSL}_2 (\mathbb{R})$ with signature $(1; \infty)$.
Shinichi Mochizuki gave a purely geometric characterization of the
corresponding arithmetic $(1; \infty)$-curves, which also arise naturally in
the context of his recent work on inter-universal Teichmüller theory.
Using Bely\u{\i} maps, we explicitly determine the canonical models of these
curves. We also study their arithmetic properties and modular interpretations. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: State observation and sensor selection for nonlinear networks,
Abstract: A large variety of dynamical systems, such as chemical and biomolecular
systems, can be seen as networks of nonlinear entities. Prediction, control,
and identification of such nonlinear networks require knowledge of the state of
the system. However, network states are usually unknown, and only a fraction of
the state variables are directly measurable. The observability problem concerns
reconstructing the network state from this limited information. Here, we
propose a general optimization-based approach for observing the states of
nonlinear networks and for optimally selecting the observed variables. Our
results reveal several fundamental limitations in network observability, such
as the trade-off between the fraction of observed variables and the observation
length on one side, and the estimation error on the other side. We also show
that owing to the crucial role played by the dynamics, purely graph- theoretic
observability approaches cannot provide conclusions about one's practical
ability to estimate the states. We demonstrate the effectiveness of our methods
by finding the key components in biological and combustion reaction networks
from which we determine the full system state. Our results can lead to the
design of novel sensing principles that can greatly advance prediction and
control of the dynamics of such networks. | [
1,
0,
1,
0,
0,
0
] | [
"Computer Science",
"Mathematics",
"Quantitative Biology"
] |
Title: Maximal fluctuations of confined actomyosin gels: dynamics of the cell nucleus,
Abstract: We investigate the effect of stress fluctuations on the stochastic dynamics
of an inclusion embedded in a viscous gel. We show that, in non-equilibrium
systems, stress fluctuations give rise to an effective attraction towards the
boundaries of the confining domain, which is reminiscent of an active Casimir
effect. We apply this generic result to the dynamics of deformations of the
cell nucleus and we demonstrate the appearance of a fluctuation maximum at a
critical level of activity, in agreement with recent experiments [E. Makhija,
D. S. Jokhun, and G. V. Shivashankar, Proc. Natl. Acad. Sci. U.S.A. 113, E32
(2016)]. | [
0,
1,
0,
0,
0,
0
] | [
"Quantitative Biology",
"Physics"
] |
Title: Information Planning for Text Data,
Abstract: Information planning enables faster learning with fewer training examples. It
is particularly applicable when training examples are costly to obtain. This
work examines the advantages of information planning for text data by focusing
on three supervised models: Naive Bayes, supervised LDA and deep neural
networks. We show that planning based on entropy and mutual information
outperforms random selection baseline and therefore accelerates learning. | [
0,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: A character of Siegel modular group of level 2 from theta constants,
Abstract: Given a characteristic, we define a character of the Siegel modular group of
level 2, the computations of their values are also obtained. By using our
theorems, some key theorems of Igusa [1] can be recovered. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics"
] |
Title: Antiferromagnetic Chern insulators in non-centrosymmetric systems,
Abstract: We investigate a new class of topological antiferromagnetic (AF) Chern
insulators driven by electronic interactions in two-dimensional systems without
inversion symmetry. Despite the absence of a net magnetization, AF Chern
insulators (AFCI) possess a nonzero Chern number $C$ and exhibit the quantum
anomalous Hall effect (QAHE). Their existence is guaranteed by the bifurcation
of the boundary line of Weyl points between a quantum spin Hall insulator and a
topologically trivial phase with the emergence of AF long-range order. As a
concrete example, we study the phase structure of the honeycomb lattice
Kane-Mele model as a function of the inversion-breaking ionic potential and the
Hubbard interaction. We find an easy $z$-axis $C=1$ AFCI phase and a spin-flop
transition to a topologically trivial $xy$-plane collinear antiferromagnet. We
propose experimental realizations of the AFCI and QAHE in correlated electron
materials and cold atom systems. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Effects of the structural distortion on the electronic band structure of {\boldmath $\rm Na Os O_3$} studied within density functional theory and a three-orbital model,
Abstract: Effects of the structural distortion associated with the $\rm OsO_6$
octahedral rotation and tilting on the electronic band structure and magnetic
anisotropy energy for the $5d^3$ compound NaOsO$_3$ are investigated using the
density functional theory (DFT) and within a three-orbital model. Comparison of
the essential features of the DFT band structures with the three-orbital model
for both the undistorted and distorted structures provides insight into the
orbital and directional asymmetry in the electron hopping terms resulting from
the structural distortion. The orbital mixing terms obtained in the transformed
hopping Hamiltonian resulting from the octahedral rotations are shown to
account for the fine features in the DFT band structure. Staggered
magnetization and the magnetic character of states near the Fermi energy
indicate weak coupling behavior. | [
0,
1,
0,
0,
0,
0
] | [
"Physics"
] |
Title: Learning What Data to Learn,
Abstract: Machine learning is essentially the sciences of playing with data. An
adaptive data selection strategy, enabling to dynamically choose different data
at various training stages, can reach a more effective model in a more
efficient way. In this paper, we propose a deep reinforcement learning
framework, which we call \emph{\textbf{N}eural \textbf{D}ata \textbf{F}ilter}
(\textbf{NDF}), to explore automatic and adaptive data selection in the
training process. In particular, NDF takes advantage of a deep neural network
to adaptively select and filter important data instances from a sequential
stream of training data, such that the future accumulative reward (e.g., the
convergence speed) is maximized. In contrast to previous studies in data
selection that is mainly based on heuristic strategies, NDF is quite generic
and thus can be widely suitable for many machine learning tasks. Taking neural
network training with stochastic gradient descent (SGD) as an example,
comprehensive experiments with respect to various neural network modeling
(e.g., multi-layer perceptron networks, convolutional neural networks and
recurrent neural networks) and several applications (e.g., image classification
and text understanding) demonstrate that NDF powered SGD can achieve comparable
accuracy with standard SGD process by using less data and fewer iterations. | [
1,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: Zero-Shot Learning via Class-Conditioned Deep Generative Models,
Abstract: We present a deep generative model for learning to predict classes not seen
at training time. Unlike most existing methods for this problem, that represent
each class as a point (via a semantic embedding), we represent each seen/unseen
class using a class-specific latent-space distribution, conditioned on class
attributes. We use these latent-space distributions as a prior for a supervised
variational autoencoder (VAE), which also facilitates learning highly
discriminative feature representations for the inputs. The entire framework is
learned end-to-end using only the seen-class training data. The model infers
corresponding attributes of a test image by maximizing the VAE lower bound; the
inferred attributes may be linked to labels not seen when training. We further
extend our model to a (1) semi-supervised/transductive setting by leveraging
unlabeled unseen-class data via an unsupervised learning module, and (2)
few-shot learning where we also have a small number of labeled inputs from the
unseen classes. We compare our model with several state-of-the-art methods
through a comprehensive set of experiments on a variety of benchmark data sets. | [
1,
0,
0,
0,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Title: On the vanishing viscosity approximation of a nonlinear model for tumor growth,
Abstract: We investigate the dynamics of a nonlinear system modeling tumor growth with
drug application. The tumor is viewed as a mixture consisting of proliferating,
quiescent and dead cells as well as a nutrient in the presence of a drug. The
system is given by a multi-phase flow model: the densities of the different
cells are governed by a set of transport equations, the density of the nutrient
and the density of the drug are governed by rather general diffusion equations,
while the velocity of the tumor is given by Darcy's equation. The domain
occupied by the tumor in this setting is a growing continuum $\Omega$ with
boundary $\partial \Omega$ both of which evolve in time. Global-in-time weak
solutions are obtained using an approach based on the vanishing viscosity of
the Brinkman's regularization. Both the solutions and the domain are rather
general, no symmetry assumption is required and the result holds for large
initial data. | [
0,
0,
1,
0,
0,
0
] | [
"Mathematics",
"Quantitative Biology"
] |
Title: Variational Bayesian dropout: pitfalls and fixes,
Abstract: Dropout, a stochastic regularisation technique for training of neural
networks, has recently been reinterpreted as a specific type of approximate
inference algorithm for Bayesian neural networks. The main contribution of the
reinterpretation is in providing a theoretical framework useful for analysing
and extending the algorithm. We show that the proposed framework suffers from
several issues; from undefined or pathological behaviour of the true posterior
related to use of improper priors, to an ill-defined variational objective due
to singularity of the approximating distribution relative to the true
posterior. Our analysis of the improper log uniform prior used in variational
Gaussian dropout suggests the pathologies are generally irredeemable, and that
the algorithm still works only because the variational formulation annuls some
of the pathologies. To address the singularity issue, we proffer Quasi-KL (QKL)
divergence, a new approximate inference objective for approximation of
high-dimensional distributions. We show that motivations for variational
Bernoulli dropout based on discretisation and noise have QKL as a limit.
Properties of QKL are studied both theoretically and on a simple practical
example which shows that the QKL-optimal approximation of a full rank Gaussian
with a degenerate one naturally leads to the Principal Component Analysis
solution. | [
0,
0,
0,
1,
0,
0
] | [
"Computer Science",
"Statistics"
] |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.