abstract
stringlengths 42
2.09k
|
|---|
Mel-filterbanks are fixed, engineered audio features which emulate human
perception and have been used through the history of audio understanding up to
today. However, their undeniable qualities are counterbalanced by the
fundamental limitations of handmade representations. In this work we show that
we can train a single learnable frontend that outperforms mel-filterbanks on a
wide range of audio signals, including speech, music, audio events and animal
sounds, providing a general-purpose learned frontend for audio classification.
To do so, we introduce a new principled, lightweight, fully learnable
architecture that can be used as a drop-in replacement of mel-filterbanks. Our
system learns all operations of audio features extraction, from filtering to
pooling, compression and normalization, and can be integrated into any neural
network at a negligible parameter cost. We perform multi-task training on eight
diverse audio classification tasks, and show consistent improvements of our
model over mel-filterbanks and previous learnable alternatives. Moreover, our
system outperforms the current state-of-the-art learnable frontend on Audioset,
with orders of magnitude fewer parameters.
|
With the dramatic rise in high-quality galaxy data expected from Euclid and
Vera C. Rubin Observatory, there will be increasing demand for fast
high-precision methods for measuring galaxy fluxes. These will be essential for
inferring the redshifts of the galaxies. In this paper, we introduce Lumos, a
deep learning method to measure photometry from galaxy images. Lumos builds on
BKGnet, an algorithm to predict the background and its associated error, and
predicts the background-subtracted flux probability density function. We have
developed Lumos for data from the Physics of the Accelerating Universe Survey
(PAUS), an imaging survey using 40 narrow-band filter camera (PAUCam). PAUCam
images are affected by scattered light, displaying a background noise pattern
that can be predicted and corrected for. On average, Lumos increases the SNR of
the observations by a factor of 2 compared to an aperture photometry algorithm.
It also incorporates other advantages like robustness towards distorting
artifacts, e.g. cosmic rays or scattered light, the ability of deblending and
less sensitivity to uncertainties in the galaxy profile parameters used to
infer the photometry. Indeed, the number of flagged photometry outlier
observations is reduced from 10% to 2%, comparing to aperture photometry.
Furthermore, with Lumos photometry, the photo-z scatter is reduced by ~10% with
the Deepz machine learning photo-z code and the photo-z outlier rate by 20%.
The photo-z improvement is lower than expected from the SNR increment, however
currently the photometric calibration and outliers in the photometry seem to be
its limiting factor.
|
It is well known that quantum effects may lead to remove the intrinsic
singularity point of back holes. Also, the quintessence scalar field is a
candidate model for describing late-time acceleration expansion. Accordingly,
Kazakov and Solodukhin considered the existence of back-reaction of the
spacetime due to the quantum fluctuations of the background metric to deform
Schwarzschild black hole, which led to change the intrinsic singularity of the
black hole to a 2-sphere with a radius of the order of the Planck length. Also,
Kiselev rewrote the Schwarzschild metric by taking into account the
quintessence field in the background. In this study, we consider the
quantum-corrected Schwarzschild black hole inspired by Kazakov-Solodukhin's
work, and Schwarzschild black hole surrounded by quintessence deduced by
Kiselev to study the mutual effects of quantum fluctuations and quintessence on
the accretion onto the black hole. Consequently, the radial component of
4-velocity and the proper energy density of the accreting fluid have a finite
value on the surface of its central 2-sphere due to the presence of quantum
corrections. Also, by comparing the accretion parameters in different kinds of
black holes, we infer that the presence of a point-like electric charge in the
spacetime is somewhat similar to some quantum fluctuations in the background
metric.
|
Batched network coding is a variation of random linear network coding which
has low computational and storage costs. In order to adapt to random
fluctuations in the number of erasures in individual batches, it is not optimal
to recode and transmit the same number of packets for all batches. Different
distributed optimization models, which are called adaptive recoding schemes,
were formulated for this purpose. The key component of these optimization
problems is the expected value of the rank distribution of a batch at the next
network node, which is also known as the expected rank. In this paper, we put
forth a unified adaptive recoding framework with an arbitrary recoding field
size. We show that the expected rank functions are concave when the packet loss
pattern is a stationary stochastic process, which covers but not limited to
independent packet loss and Gilbert-Elliott packet loss model. Under this
concavity assumption, we show that there always exists a solution which not
only can minimize the randomness on the number of recoded packets but also can
tolerate rank distribution errors due to inaccurate measurements or limited
precision of the machine. We provide an algorithm to obtain such an optimal
optimal solution, and propose tuning schemes that can turn any feasible
solution into a desired optimal solution.
|
Based on direct numerical simulations with point-like inertial particles
transported by homogeneous and isotropic turbulent flows, we present evidence
for the existence of Markov property in Lagrangian turbulence. We show that the
Markov property is valid for a finite step size larger than a Stokes
number-dependent Einstein-Markov memory length. This enables the description of
multi-scale statistics of Lagrangian particles by Fokker-Planck equations,
which can be embedded in an interdisciplinary approach linking the statistical
description of turbulence with fluctuation theorems of non-equilibrium
stochastic thermodynamics and fluctuation theorems, and local flow structures.
|
A dedicated in situ heating setup in a scanning electron microscope (SEM)
followed by an ex situ atomic force microscopy (AFM) and electron backscatter
diffraction (EBSD) is used to characterize the nucleation and early growth
stages of Fe-Al intermetallics (IMs) at 596 {\deg}C. A location tracking is
used to interpret further characterization. Ex situ AFM observations reveal a
slight shrinkage and out of plane protrusion of the IM at the onset of IM
nucleation followed by directional growth. The formed interfacial IM compounds
were identified by ex situ EBSD. It is now clearly demonstrated that the
{\theta}-phase nucleates first prior to the diffusion-controlled growth of the
{\eta}-phase. The {\theta}-phase prevails the intermetallic layer.
|
We present a conceptual study of a large format imaging spectrograph for the
Large Submillimeter Telescope (LST) and the Atacama Large Aperture
Submillimeter Telescope (AtLAST). Recent observations of high-redshift galaxies
indicate the onset of earliest star formation just a few 100 million years
after the Big Bang (i.e., z = 12--15), and LST/AtLAST will provide a unique
pathway to uncover spectroscopically-identified first forming galaxies in the
pre-reionization era, once it will be equipped with a large format imaging
spectrograph. We propose a 3-band (200, 255, and 350 GHz), medium resolution (R
= 2,000) imaging spectrograph with 1.5 M detectors in total based on the KATANA
concept (Karatsu et al. 2019), which exploits technologies of the integrated
superconducting spectrometer (ISS) and a large-format imaging array. A 1-deg2
drilling survey (3,500 hr) will capture a large number of [O III] 88 um (and [C
II] 158 um) emitters at z = 8--9, and constrain [O III] luminosity functions at
z > 12.
|
Score-based diffusion models synthesize samples by reversing a stochastic
process that diffuses data to noise, and are trained by minimizing a weighted
combination of score matching losses. The log-likelihood of score-based
diffusion models can be tractably computed through a connection to continuous
normalizing flows, but log-likelihood is not directly optimized by the weighted
combination of score matching losses. We show that for a specific weighting
scheme, the objective upper bounds the negative log-likelihood, thus enabling
approximate maximum likelihood training of score-based diffusion models. We
empirically observe that maximum likelihood training consistently improves the
likelihood of score-based diffusion models across multiple datasets, stochastic
processes, and model architectures. Our best models achieve negative
log-likelihoods of 2.83 and 3.76 bits/dim on CIFAR-10 and ImageNet 32x32
without any data augmentation, on a par with state-of-the-art autoregressive
models on these tasks.
|
We study a model system with nematic and magnetic orders, within a channel
geometry modelled by an interval, $[-D, D]$. The system is characterised by a
tensor-valued nematic order parameter $\mathbf{Q}$ and a vector-valued
magnetisation $\mathbf{M}$, and the observable states are modelled as stable
critical points of an appropriately defined free energy. In particular, the
full energy includes a nemato-magnetic coupling term characterised by a
parameter $c$. We (i) derive $L^\infty$ bounds for $\mathbf{Q}$ and
$\mathbf{M}$; (ii) prove a uniqueness result in parameter regimes defined by
$c$, $D$ and material- and temperature-dependent correlation lengths; (iii)
analyse order reconstruction solutions, possessing domain walls, and their
stabilities as a function of $D$ and $c$ and (iv) perform numerical studies
that elucidate the interplay of $c$ and $D$ for multistability.
|
We propose a mechanism to substantially rectify radiative heat flow by
matching thin films of metal-to-insulator transition materials and polar
dielectrics in the electromagnetic near field. By leveraging the distinct
scaling behaviors of the local density of states with film thickness for metals
and insulators, we theoretically achieve rectification ratios over 140-a
10-fold improvement over the state of the art-with nanofilms of vanadium
dioxide and cubic boron nitride in the parallel-plane geometry at
experimentally feasible gap sizes (~100 nm). Our rational design offers
relative ease of fabrication, flexible choice of materials, and robustness
against deviations from optimal film thicknesses. We expect this work to
facilitate the application of thermal diodes in solid-state thermal circuits
and energy conversion devices.
|
We propose a predictive model of the turbulent burning velocity over a wide
range of conditions. The model consists of sub models of the stretch factor and
the turbulent flame area. The stretch factor characterizes the flame response
of turbulence stretch and incorporates effects of detailed chemistry and
transport with a lookup table of laminar counterflow flames. The flame area
model captures the area growth based on Lagrangian statistics of propagating
surfaces, and considers effects of turbulence length scales and fuel
characteristics. The present model predicts the turbulent burning velocity via
an algebraic expression without free parameters. It is validated against 285
cases of the direct numerical simulation or experiment reported from various
research groups on planar and Bunsen flames over a wide range of conditions,
covering fuels from hydrogen to n-dodecane, pressures from 1 to 20 atm, lean
and rich mixtures, turbulence intensity ratios from 0.35 to 110, and turbulence
length ratios from 0.5 to 80. The comprehensive comparison shows that the
proposed turbulent burning velocity model has an overall good agreement over
the wide range of conditions, with the averaged modeling error of 25.3%.
Furthermore, the model prediction involves the uncertainty quantification for
model parameters and chemical kinetics to extend the model applicability.
|
We propose a distributed approach to train deep convolutional generative
adversarial neural network (DC-CGANs) models. Our method reduces the imbalance
between generator and discriminator by partitioning the training data according
to data labels, and enhances scalability by performing a parallel training
where multiple generators are concurrently trained, each one of them focusing
on a single data label. Performance is assessed in terms of inception score and
image quality on MNIST, CIFAR10, CIFAR100, and ImageNet1k datasets, showing a
significant improvement in comparison to state-of-the-art techniques to
training DC-CGANs. Weak scaling is attained on all the four datasets using up
to 1,000 processes and 2,000 NVIDIA V100 GPUs on the OLCF supercomputer Summit.
|
In this paper we show how inter-cellular molecular communication may change
the overall levels of photosynthesis in plants. Individual plant cells respond
to external stimuli, such as illumination levels, to regulate their
photosynthetic output. Here, we present a mathematical model which shows that
by sharing information internally using molecular communication, plants may
increase overall photosynthate production. Numerical results show that higher
mutual information between cells corresponds to an increase in overall
photosynthesis by as much as 25 per cent. This suggests that molecular
communication plays a vital role in maximising the photosynthesis in plants and
therefore suggests new routes to influence plant development in agriculture and
elsewhere.
|
In this paper, the split common null point problem in two Banach spaces is
considered. Then, using the generalized resolvents of maximal monotone
operators and the generalized projections and an infinite family of
nonexpansive mappings, a strong convergence theorem for finding a solution of
the split common null point problem in two Banach spaces in the presence of a
sequence of errors will be proved.
|
Spectrally-efficient secure non-orthogonal multiple access (NOMA) has
recently attained a substantial research interest for fifth generation
development. This work explores crucial security issue in NOMA which is stemmed
from utilizing the decoding concept of successive interference cancellation.
Considering untrusted users, we design a novel secure NOMA transmission
protocol to maximize secrecy fairness among users. A new decoding order for two
users' NOMA is proposed that provides positive secrecy rate to both users.
Observing the objective of maximizing secrecy fairness between users under
given power budget constraint, the problem is formulated as minimizing the
maximum secrecy outage probability (SOP) between users. In particular,
closed-form expressions of SOP for both users are derived to analyze secrecy
performance. SOP minimization problems are solved using pseudoconvexity
concept, and optimized power allocation (PA) for each user is obtained.
Asymptotic expressions of SOPs, and optimal PAs minimizing these approximations
are obtained to get deeper insights. Further, globally-optimized power control
solution from secrecy fairness perspective is obtained at a low computational
complexity and, asymptotic approximation is obtained to gain analytical
insights. Numerical results validate the correctness of analysis, and present
insights on optimal solutions. Finally, we present insights on global-optimal
PA by which fairness is ensured and gains of about 55.12%, 69.30%, and 19.11%,
respectively are achieved, compared to fixed PA and individual users' optimal
PAs.
|
Variational inference enables approximate posterior inference of the highly
over-parameterized neural networks that are popular in modern machine learning.
Unfortunately, such posteriors are known to exhibit various pathological
behaviors. We prove that as the number of hidden units in a single-layer
Bayesian neural network tends to infinity, the function-space posterior mean
under mean-field variational inference actually converges to zero, completely
ignoring the data. This is in contrast to the true posterior, which converges
to a Gaussian process. Our work provides insight into the over-regularization
of the KL divergence in variational inference.
|
We present TransitFit, an open-source Python~3 package designed to fit
exoplanetary transit light-curves for transmission spectroscopy studies
(Available at https://github.com/joshjchayes/TransitFit and
https://github.com/spearnet/TransitFit, with documentation at
https://transitfit.readthedocs.io/). TransitFit employs nested sampling to
offer efficient and robust multi-epoch, multi-wavelength fitting of transit
data obtained from one or more telescopes. TransitFit allows per-telescope
detrending to be performed simultaneously with parameter fitting, including the
use of user-supplied detrending alogorithms. Host limb darkening can be fitted
either independently ("uncoupled") for each filter or combined ("coupled")
using prior conditioning from the PHOENIX stellar atmosphere models. For this
TransitFit uses the Limb Darkening Toolkit (LDTk) together with filter
profiles, including user-supplied filter profiles. We demonstrate the
application of TransitFit in three different contexts. First, we model SPEARNET
broadband optical data of the low-density hot-Neptune WASP-127~b. The data were
obtained from a globally-distributed network of 0.5m--2.4m telescopes. We find
clear improvement in our broadband results using the coupled mode over
uncoupled mode, when compared against the higher spectral resolution GTC/OSIRIS
transmission spectrum obtained by Chen et al. (2018). Using TransitFit, we fit
26 transit observations by TESS to recover improved ephemerides of the
hot-Jupiter WASP-91~b and a transit depth determined to a precision of 170~ppm.
Finally, we use TransitFit to conduct an investigation into the contested
presence of TTV signatures in WASP-126~b using 126 transits observed by TESS,
concluding that there is no statistically significant evidence for such
signatures from observations spanning 31 TESS sectors.
|
The ongoing Coronavirus disease 2019 (COVID-19) is a major crisis that has
significantly affected the healthcare sector and global economies, which made
it the main subject of various fields in scientific and technical research. To
properly understand and control this new epidemic, mathematical modelling is
presented as a very effective tool that can illustrate the mechanisms of its
propagation. In this regard, the use of compartmental models is the most
prominent approach adopted in the literature to describe the dynamics of
COVID-19. Along the same line, we aim during this study to generalize and
ameliorate many existing works that consecrated to analyse the behaviour of
this epidemic. Precisely, we propose an SQEAIHR epidemic system for
Coronavirus. Our constructed model is enriched by taking into account the media
intervention and vital dynamics. By the use of the next-generation matrix
method, the theoretical basic reproductive number $R_0$ is obtained for
COVID-19. Based on some nonstandard and generalized analytical techniques, the
local and global stability of the disease-free equilibrium are proven when $R_0
< 1$. Moreover, in the case of $R_0 > 1$, the uniform persistence of COVID-19
model is also shown. In order to better adapt our epidemic model to reality,
the randomness factor is taken into account by considering a proportional white
noises, which leads to a well-posed stochastic model. Under appropriate
conditions, interesting asymptotic properties are proved, namely: extinction
and persistence in the mean. The theoretical results show that the dynamics of
the perturbed COVID-19 model are determined by parameters that are closely
related to the magnitude of the stochastic noise. Finally, we present some
numerical illustrations to confirm our theoretical results and to show the
impact of media intervention and quarantine strategies.
|
In this work, we generalize the reaction-diffusion equation in statistical
physics, Schr\"odinger equation in quantum mechanics, Helmholtz equation in
paraxial optics into the neural partial differential equations (NPDE), which
can be considered as the fundamental equations in the field of artificial
intelligence research. We take finite difference method to discretize NPDE for
finding numerical solution, and the basic building blocks of deep neural
network architecture, including multi-layer perceptron, convolutional neural
network and recurrent neural networks, are generated. The learning strategies,
such as Adaptive moment estimation, L-BFGS, pseudoinverse learning algorithms
and partial differential equation constrained optimization, are also presented.
We believe it is of significance that presented clear physical image of
interpretable deep neural networks, which makes it be possible for applying to
analog computing device design, and pave the road to physical artificial
intelligence.
|
Generative modeling has recently shown great promise in computer vision, but
it has mostly focused on synthesizing visually realistic images. In this paper,
motivated by multi-task learning of shareable feature representations, we
consider a novel problem of learning a shared generative model that is useful
across various visual perception tasks. Correspondingly, we propose a general
multi-task oriented generative modeling (MGM) framework, by coupling a
discriminative multi-task network with a generative network. While it is
challenging to synthesize both RGB images and pixel-level annotations in
multi-task scenarios, our framework enables us to use synthesized images paired
with only weak annotations (i.e., image-level scene labels) to facilitate
multiple visual tasks. Experimental evaluation on challenging multi-task
benchmarks, including NYUv2 and Taskonomy, demonstrates that our MGM framework
improves the performance of all the tasks by large margins, consistently
outperforming state-of-the-art multi-task approaches.
|
Understanding the physics of strongly correlated electronic systems has been
a central issue in condensed matter physics for decades. In transition metal
oxides, strong correlations characteristic of narrow $d$ bands is at the origin
of such remarkable properties as the Mott gap opening, enhanced effective mass,
and anomalous vibronic coupling, to mention a few. SrVO$_3$, with V$^{4+}$ in a
$3d^1$ electronic configuration is the simplest example of a 3D correlated
metallic electronic system. Here, we focus on the observation of a (roughly)
quadratic temperature dependence of the inverse electron mobility of this
seemingly simple system, which is an intriguing property shared by other
metallic oxides. The systematic analysis of electronic transport in SrVO$_3$
thin films discloses the limitations of the simplest picture of e-e
correlations in a Fermi liquid; instead, we show that the quasi-2D topology of
the Fermi surface and a strong electron-phonon coupling, contributing to dress
carriers with a phonon cloud, play a pivotal role on the reported electron
spectroscopic, optical, thermodynamic and transport data. The picture that
emerges is not restricted to SrVO$_3$ but can be shared with other $3d$ and
$4d$ metallic oxides.
|
Let $ E \subset \mathbb{R}^2 $ be a finite set, and let $ f : E \to
[0,\infty) $. In this paper, we address the algorithmic aspects of nonnegative
$C^2$ interpolation in the plane. Specifically, we provide an efficient
algorithm to compute a nonnegative $C^2(\mathbb{R}^2)$ extension of $ f $ with
norm within a universal constant factor of the least possible. We also provide
an efficient algorithm to approximate the trace norm.
|
In the past decades, the revolutionary advances of Machine Learning (ML) have
shown a rapid adoption of ML models into software systems of diverse types.
Such Machine Learning Software Applications (MLSAs) are gaining importance in
our daily lives. As such, the Quality Assurance (QA) of MLSAs is of paramount
importance. Several research efforts are dedicated to determining the specific
challenges we can face while adopting ML models into software systems. However,
we are aware of no research that offered a holistic view of the distribution of
those ML quality assurance challenges across the various phases of software
development life cycles (SDLC). This paper conducts an in-depth literature
review of a large volume of research papers that focused on the quality
assurance of ML models. We developed a taxonomy of MLSA quality assurance
issues by mapping the various ML adoption challenges across different phases of
SDLC. We provide recommendations and research opportunities to improve SDLC
practices based on the taxonomy. This mapping can help prioritize quality
assurance efforts of MLSAs where the adoption of ML models can be considered
crucial.
|
Giant Radio Galaxies (GRGs) are the largest single structures in the
Universe. Exhibiting extended radio morphology, their projected sizes range
from 0.7 Mpc up to 4.9 Mpc. LOFAR has opened a new window on the discovery and
investigation of GRGs and, despite the hundreds that are today known, their
main growth catalyst is still debated. One natural explanation for the
exceptional size of GRGs is their old age. In this context, hard X-ray selected
GRGs show evidence of restarting activity, with the giant radio lobes being
mostly disconnected from the nuclear source, if any. In this paper, we present
the serendipitous discovery of a distant ($z=0.629$), medium X-ray selected GRG
in the Bo\"otes field. High-quality, deep Chandra and LOFAR data allow a robust
study of the connection between the nucleus and the lobes, at a larger redshift
so far inaccessible to coded-mask hard X-ray instruments. The radio morphology
of the GRG presented in this work does not show evidence for restarted
activity, and the nuclear radio core spectrum does not appear to be GPS-like.
On the other hand, the X-ray properties of the new GRG are perfectly consistent
with the ones previously studied with Swift/BAT and INTEGRAL at lower redshift.
In particular, the bolometric luminosity measured from the X-ray spectrum is a
factor of six larger than the one derived from the radio lobes, although the
large uncertainties make them formally consistent at $1\sigma$. Finally, the
moderately dense environment around the GRG, traced by the spatial distribution
of galaxies, supports recent findings that the growth of GRGs is not primarily
driven by underdense environments.
|
By employing a pseudo-orthonormal coordinate-free approach, the Dirac
equation for particles in the Kerr--Newman spacetime is separated into its
radial and angular parts. In the massless case to which a special attention is
given, the general Heun-type equations turn into their confluent form. We show
how one recovers some results previously obtained in literature, by other
means.
|
Although many techniques have been applied to matrix factorization (MF), they
may not fully exploit the feature structure. In this paper, we incorporate the
grouping effect into MF and propose a novel method called Robust Matrix
Factorization with Grouping effect (GRMF). The grouping effect is a
generalization of the sparsity effect, which conducts denoising by clustering
similar values around multiple centers instead of just around 0. Compared with
existing algorithms, the proposed GRMF can automatically learn the grouping
structure and sparsity in MF without prior knowledge, by introducing a
naturally adjustable non-convex regularization to achieve simultaneous sparsity
and grouping effect. Specifically, GRMF uses an efficient alternating
minimization framework to perform MF, in which the original non-convex problem
is first converted into a convex problem through Difference-of-Convex (DC)
programming, and then solved by Alternating Direction Method of Multipliers
(ADMM). In addition, GRMF can be easily extended to the Non-negative Matrix
Factorization (NMF) settings. Extensive experiments have been conducted using
real-world data sets with outliers and contaminated noise, where the
experimental results show that GRMF has promoted performance and robustness,
compared to five benchmark algorithms.
|
The increasing market penetration of electric vehicles (EVs) may pose
significant electricity demand on power systems. This electricity demand is
affected by the inherent uncertainties of EVs' travel behavior that makes
forecasting the daily charging demand (CD) very challenging. In this project,
we use the National House Hold Survey (NHTS) data to form sequences of trips,
and develop machine learning models to predict the parameters of the next trip
of the drivers, including trip start time, end time, and distance. These
parameters are later used to model the temporal charging behavior of EVs. The
simulation results show that the proposed modeling can effectively estimate the
daily CD pattern based on travel behavior of EVs, and simple machine learning
techniques can forecast the travel parameters with acceptable accuracy.
|
We argue that neutrino oscillations at JUNO offer a unique opportunity to
study Sorkin's triple-path interference, which is predicted to be zero in
canonical quantum mechanics by virtue of the Born rule. In particular, we
compute the expected bounds on triple-path interference at JUNO and demonstrate
that they are comparable to those already available from electromagnetic
probes. Furthermore, the neutrino probe of the Born rule is much more direct
due to an intrinsic independence from any boundary conditions, whereas such
dependence on boundary conditions is always present in the case of
electromagnetic probes. Thus, neutrino oscillations present an ideal probe of
this aspect of the foundations of quantum mechanics.
|
This paper is concerned with the asymptotic analysis of sojourn times of
random fields with continuous sample paths. Under a very general framework we
show that there is an interesting relationship between tail asymptotics of
sojourn times and that of supremum. Moreover, we establish the uniform
double-sum method to derive the tail asymptotics of sojourn times. In the
literature, based on the pioneering research of S. Berman the sojourn times
have been utilised to derive the tail asymptotics of supremum of Gaussian
processes. In this paper we show that the opposite direction is even more
fruitful, namely knowing the asymptotics of supremum o f random processes and
fields (in particular Gaussian) it is possible to establish the asymptotics of
their sojourn times. We illustrate our findings considering i) two dimensional
Gaussian random fields, ii) chi-process generated by stationary Gaussian
processes and iii) stationary Gaussian queueing processes.
|
We consider the setting where the nodes of an undirected, connected network
collaborate to solve a shared objective modeled as the sum of smooth functions.
We assume that each summand is privately known by a unique node. NEAR-DGD is a
distributed first order method which permits adjusting the amount of
communication between nodes relative to the amount of computation performed
locally in order to balance convergence accuracy and total application cost. In
this work, we generalize the convergence properties of a variant of NEAR-DGD
from the strongly convex to the nonconvex case. Under mild assumptions, we show
convergence to minimizers of a custom Lyapunov function. Moreover, we
demonstrate that the gap between those minimizers and the second order
stationary solutions of the original problem can become arbitrarily small
depending on the choice of algorithm parameters. Finally, we accompany our
theoretical analysis with a numerical experiment to evaluate the empirical
performance of NEAR-DGD in the nonconvex setting.
|
We calculate models of stellar evolution for very massive stars and include
the effects of modified gravity to investigate the influence on the physical
properties of blue supergiant stars and their use as extragalactic distance
indicators. With shielding and fifth force parameters in a similar range as in
previous studies of Cepheid and tip of the red giant branch (TRGB) stars we
find clear effects on stellar luminosity and flux-weighted gravity. The
relationship between flux weighted gravity, g_F = g/Teff^4, and bolometric
magnitude M_bol (FGLR), which has been used successfully for accurate distance
determinations, is systematically affected. While the stellar evolution FGLRs
show a systematic offset from the observed relation, we can use the
differential shifts between models with Newtonian and modified gravity to
estimate the influence on FGLR distance determinations. Modified gravity leads
to a distance increase of 0.05 to 0.15 magnitudes in distance modulus. These
change are comparable to the ones found for Cepheid stars. We compare observed
FGLR and TRGB distances of nine galaxies to constrain the free parameters of
modified gravity. Not accounting for systematic differences between TRGB and
FGLR distances shielding parameters of 5*10^-7 and 10^-6 and fifth force
parameters of 1/3 and 1 can be ruled out with about 90% confidence. Allowing
for potential systematic offsets between TRGB and FGLR distances no
determination is possible for a shielding parameter of 10^-6. For 5*10$^-7 a
fifth force parameter of 1 can be ruled out to 92% but 1/3 is unlikely only to
60%.
|
The superior performance of CNN on medical image analysis heavily depends on
the annotation quality, such as the number of labeled image, the source of
image, and the expert experience. The annotation requires great expertise and
labour. To deal with the high inter-rater variability, the study of imperfect
label has great significance in medical image segmentation tasks. In this
paper, we present a novel cascaded robust learning framework for chest X-ray
segmentation with imperfect annotation. Our model consists of three independent
network, which can effectively learn useful information from the peer networks.
The framework includes two stages. In the first stage, we select the clean
annotated samples via a model committee setting, the networks are trained by
minimizing a segmentation loss using the selected clean samples. In the second
stage, we design a joint optimization framework with label correction to
gradually correct the wrong annotation and improve the network performance. We
conduct experiments on the public chest X-ray image datasets collected by
Shenzhen Hospital. The results show that our methods could achieve a
significant improvement on the accuracy in segmentation tasks compared to the
previous methods.
|
The huge amount of data produced in the fifth-generation (5G) networks not
only brings new challenges to the reliability and efficiency of mobile devices
but also drives rapid development of new storage techniques. With the benefits
of fast access speed and high reliability, NAND flash memory has become a
promising storage solution for the 5G networks. In this paper, we investigate a
protograph-coded bit-interleaved coded modulation with iterative detection and
decoding (BICM-ID) utilizing irregular mapping (IM) in the multi-level-cell
(MLC) NAND flash-memory systems. First, we propose an enhanced protograph-based
extrinsic information transfer (EPEXIT) algorithm to facilitate the analysis of
protograph codes in the IM-BICM-ID systems. With the use of EPEXIT algorithm, a
simple design method is conceived for the construction of a family of high-rate
protograph codes, called irregular-mapped accumulate-repeat-accumulate (IMARA)
codes, which possess both excellent decoding thresholds and
linear-minimum-distance-growth property. Furthermore, motivated by the
voltage-region iterative gain characteristics of IM-BICM-ID systems, a novel
read-voltage optimization scheme is developed to acquire accurate read-voltage
levels, thus minimizing the decoding thresholds of protograph codes.
Theoretical analyses and error-rate simulations indicate that the proposed
IMARA-aided IM-BICM-ID scheme and the proposed read-voltage optimization scheme
remarkably improve the convergence and decoding performance of flash-memory
systems. Thus, the proposed protograph-coded IM-BICM-ID flash-memory systems
can be viewed as a reliable and efficient storage solution for the
new-generation mobile networks with massive data-storage requirement.
|
We propose and experimentally demonstrate a novel interference fading
suppression method for phase-sensitive optical time domain reflectometry
(Phi-OTDR) using space-division multiplexed (SDM) pulse probes in few-mode
fiber. The SDM probes consist of multiple different modes, and three spatial
modes (LP01, LP11a and LP11b) are used in this work for proof of concept.
Firstly, the Rayleigh backscattering light of different modes is experimentally
characterized, and it turns out that the waveforms of Phi-OTDR traces of
distinct modes are all different from each other. Thanks to the spatial
difference of fading positions of distinct modes, multiple probes from
spatially multiplexed modes can be used to suppress the interference fading in
Phi-OTDR. Then, the performances of the Phi-OTDR systems using single probe and
multiple probes are evaluated and compared. Specifically, statistical analysis
shows that both fading probabilities over fiber length and time are reduced
significantly by using multiple SDM probes, which verifies the significant
performance improvement on fading suppression. The proposed novel interference
fading suppression method does not require complicated frequency or phase
modulation, which has the advantages of simplicity, good effectiveness and high
reliability.
|
Nonuniform fast Fourier transforms dominate the computational cost in many
applications including image reconstruction and signal processing. We thus
present a general-purpose GPU-based CUDA library for type 1 (nonuniform to
uniform) and type 2 (uniform to nonuniform) transforms in dimensions 2 and 3,
in single or double precision. It achieves high performance for a given
user-requested accuracy, regardless of the distribution of nonuniform points,
via cache-aware point reordering, and load-balanced blocked spreading in shared
memory. At low accuracies, this gives on-GPU throughputs around $10^9$
nonuniform points per second, and (even including host-device transfer) is
typically 4-10$\times$ faster than the latest parallel CPU code FINUFFT (at 28
threads). It is competitive with two established GPU codes, being up to
90$\times$ faster at high accuracy and/or type 1 clustered point distributions.
Finally we demonstrate a 5-12$\times$ speedup versus CPU in an X-ray
diffraction 3D iterative reconstruction task at $10^{-12}$ accuracy, observing
excellent multi-GPU weak scaling up to one rank per GPU.
|
In this paper, we discuss the properties of the generating functions of spin
Hurwitz numbers. In particular, for spin Hurwitz numbers with arbitrary
ramification profiles, we construct the weighed sums which are given by Orlov's
hypergeometric solutions of the 2-component BKP hierarchy. We derive the closed
algebraic formulas for the correlation functions associated with these
tau-functions, and under reasonable analytical assumptions we prove the loop
equations (the blobbed topological recursion). Finally, we prove a version of
topological recursion for the spin Hurwitz numbers with the spin completed
cycles (a generalized version of the Giacchetto--Kramer--Lewa\'nski
conjecture).
|
We study inverse problems for the nonlinear wave equation $\square_g u +
w(x,u, \nabla_g u) = 0$ in a Lorentzian manifold $(M,g)$ with boundary, where
$\nabla_g u$ denotes the gradient and $w(x,u, \xi)$ is smooth and quadratic in
$\xi$. Under appropriate assumptions, we show that the conformal class of the
Lorentzian metric $g$ can be recovered up to diffeomorphisms, from the
knowledge of the Neumann-to-Dirichlet map. With some additional conditions, we
can recover the metric itself up to diffeomorphisms. Moreover, we can recover
the second and third quadratic forms in the Taylor expansion of $w(x,u, \xi)$
with respect to $u$ up to null forms.
|
This article discusses the physical and kinematical characteristics of
planetary nebulae accompanying PG1159 central stars. The study is based on the
parallax and proper motion measurements recently offered by Gaia space mission.
Two approaches were used to investigate the kinematical properties of the
sample. The results revealed that most of the studied nebulae arise from
progenitor stars of mass range; $0.9-1.75$\,M$_{\odot}$. Furthermore, they tend
to live within the Galactic thick-disk and moving with an average peculiar
velocity of $61.7\pm19.2$\,km\,s$^{-1}$ at a mean vertical height of $469\pm79$
pc. The locations of the PG1159 stars on the H-R diagram indicate that they
have an average final stellar mass and evolutionary age of
$0.58\pm0.08$\,M$_{\odot}$ and $25.5\pm5.3 \rm{x}10^3$ yr, respectively. We
found a good agreement between the mean evolutionary age of the PG1159 stars
and the mean dynamical age of their companion planetary nebulae ($28.0\pm6.4
\rm{x}10^3$ yr).
|
The pentakis dodecahedron, the dual of the truncated icosahedron, consists of
60 edge-sharing triangles. It has 20 six- and 12 five-fold coordinated
vertices, with the former forming a dodecahedron, and each of the latter
connected to the vertices of one of the 12 pentagons of the dodecahedron. When
spins mounted on the vertices of the pentakis dodecahedron interact according
to the nearest-neighbor antiferromagnetic Heisenberg model, the two different
vertex types necessitate the introduction of two exchange constants. As the
relative strength of the two constants is varied the molecule interpolates
between the dodecahedron and a molecule consisting only of quadrangles. The
competition between the two exchange constants, frustration, and an external
magnetic field results in a multitude of ground-state magnetization and
susceptibility discontinuities. At the classical level the maximum is ten
magnetization and one susceptibility discontinuities when the 12 five-fold
vertices interact with the dodecahedron spins with approximately one-half the
strength of their interaction. When the two interactions are approximately
equal in strength the number of discontinuities is also maximized, with three
of the magnetization and eight of the susceptibility. At the full quantum
limit, where the magnitude of the spins equals 1/2, there can be up to three
ground-state magnetization jumps that have the total z spin component changing
by \Delta S^z=2, even though quantum fluctuations rarely allow discontinuities
of the magnetization. The full quantum case also supports a \Delta S^z=3
discontinuity. Frustration also results in nonmagnetic states inside the
singlet-triplet gap. These results make the pentakis dodecahedron the molecule
with the most discontinuous magnetic response from the quantum to the classical
level.
|
Dynamical scaling is an asymptotic property typical for the dynamics of
first-order phase transitions in physical systems and related to
self-similarity. Based on the integral-representation for the marginal
probabilities of a fractional non-homogeneous Poisson process introduced by
Leonenko et al. (2017) and generalising the standard fractional Poisson
process, we prove the dynamical scaling under fairly mild conditions. Our
result also includes the special case of the standard fractional Poisson
process.
|
The performance of superconducting radio-frequency (SRF) cavities depends on
the niobium surface condition. Recently, various heat-treatment methods have
been investigated to achieve unprecedented high quality factor (Q) and high
accelerating field (E). We report the influence of a new baking process called
furnace baking on the Q-E behavior of 1.3 GHz SRF cavities. Furnace baking is
performed as the final step of the cavity surface treatment; the cavities are
heated in a vacuum furnace for 3 h, followed by high-pressure rinsing and
radio-frequency measurement. This method is simpler and potentially more
reliable than previously reported heat-treatment methods, and it is therefore,
easier to apply to the SRF cavities. We find that the quality factor is
increased after furnace baking at temperatures ranging from 300C to 400C, while
strong decreasing the quality factor at high accelerating field is observed
after furnace baking at temperatures ranging from 600C to 800C. We find
significant differences in the surface resistance for various processing
temperatures.
|
Sound Event Detection and Audio Classification tasks are traditionally
addressed through time-frequency representations of audio signals such as
spectrograms. However, the emergence of deep neural networks as efficient
feature extractors has enabled the direct use of audio signals for
classification purposes. In this paper, we attempt to recognize musical
instruments in polyphonic audio by only feeding their raw waveforms into deep
learning models. Various recurrent and convolutional architectures
incorporating residual connections are examined and parameterized in order to
build end-to-end classi-fiers with low computational cost and only minimal
preprocessing. We obtain competitive classification scores and useful
instrument-wise insight through the IRMAS test set, utilizing a parallel
CNN-BiGRU model with multiple residual connections, while maintaining a
significantly reduced number of trainable parameters.
|
We study the topological properties of a spin-orbit coupled Hofstadter model
on the Kagome lattice. The model is time-reversal invariant and realizes a
$\mathbb{Z}_2$ topological insulator as a result of artificial gauge fields. We
develop topological arguments to describe this system showing three
inequivalent sites in a unit cell and a flat band in its energy spectrum in
addition to the topological dispersive energy bands. We show the stability of
the topological phase towards spin-flip processes and different types of
on-site potentials. In particular, we also address the situation where on-site
energies may differ inside a unit cell. Moreover, a staggered potential on the
lattice may realize topological phases for the half-filled situation. Another
interesting result is the occurrence of a topological phase for large on-site
energies. To describe topological properties of the system we use a numerical
approach based on the twisted boundary conditions and we develop a mathematical
approach, related to smooth fields.
|
Cross-flow turbines convert kinetic power in wind or water currents to
mechanical power. Unlike axial-flow turbines, the influence of geometric
parameters on turbine performance is not well-understood, in part because there
are neither generalized analytical formulations nor inexpensive, accurate
numerical models that describe their fluid dynamics. Here, we experimentally
investigate the effect of aspect ratio - the ratio of the blade span to rotor
diameter - on the performance of a straight-bladed cross-flow turbine in a
water channel. To isolate the effect of aspect ratio, all other non-dimensional
parameters are held constant, including the relative confinement, Froude
number, and Reynolds number. The coefficient of performance is found to be
invariant for the range of aspect ratios tested (0.95 - 1.63), which we ascribe
to minimal blade-support interactions for this turbine design. Finally, a
subset of experiments is repeated without controlling for the Froude number and
the coefficient of performance is found to increase, a consequence of Froude
number variation that could mistakenly be ascribed to aspect ratio. This
highlights the importance of rigorous experimental design when exploring the
effect of geometric parameters on cross-flow turbine performance.
|
After more than a decade of intense focus on automated vehicles, we are still
facing huge challenges for the vision of fully autonomous driving to become a
reality. The same "disillusionment" is true in many other domains, in which
autonomous Cyber-Physical Systems (CPS) could considerably help to overcome
societal challenges and be highly beneficial to society and individuals. Taking
the automotive domain, i.e. highly automated vehicles (HAV), as an example,
this paper sets out to summarize the major challenges that are still to
overcome for achieving safe, secure, reliable and trustworthy highly automated
resp. autonomous CPS. We constrain ourselves to technical challenges,
acknowledging the importance of (legal) regulations, certification,
standardization, ethics, and societal acceptance, to name but a few, without
delving deeper into them as this is beyond the scope of this paper. Four
challenges have been identified as being the main obstacles to realizing HAV:
Realization of continuous, post-deployment systems improvement, handling of
uncertainties and incomplete information, verification of HAV with machine
learning components, and prediction. Each of these challenges is described in
detail, including sub-challenges and, where appropriate, possible approaches to
overcome them. By working together in a common effort between industry and
academy and focusing on these challenges, the authors hope to contribute to
overcome the "disillusionment" for realizing HAV.
|
We present an integrated design to precisely measure optical frequency using
weak value amplification with a multi-mode interferometer. The technique
involves introducing a weak perturbation to the system and then post-selecting
the data in such a way that the signal is amplified without amplifying the
technical noise, as has previously been demonstrated in a free-space setup. We
demonstrate the advantages of a Bragg grating with two band gaps for obtaining
simultaneous, stable high transmission and high dispersion. We numerically
model the interferometer in order to demonstrate the amplification effect. The
device is shown to have advantages over both the free-space implementation and
other methods of measuring optical frequency on a chip, such as an integrated
Mach-Zehnder interferometer.
|
In this paper, we introduce single acceptance sampling inspection plan
(SASIP) for transmuted Rayleigh (TR) distribution when the lifetime experiment
is truncated at a prefixed time. Establish the proposed plan for different
choices of confidence level, acceptance number and ratio of true mean lifetime
to specified mean lifetime. Minimum sample size necessary to ensure a certain
specified lifetime is obtained. Operating characteristic(OC) values and
producer's risk of proposed plan are presented. Two real life example has been
presented to show the applicability of proposed SASIP.
|
In this paper, we study the periodicity structure of finite field linear
recurring sequences whose period is not necessarily maximal and determine
necessary and sufficient conditions for the characteristic polynomial~\(f\) to
have exactly two periods in the sense that the period of any sequence generated
by~\(f\) is either one or a unique integer greater than one.
|
We present a comprehensive analysis of all XMM-Newton spectra of OJ 287
spanning 15 years of X-ray spectroscopy of this bright blazar. We also report
the latest results from our dedicated Swift UVOT and XRT monitoring of OJ 287
which started in 2015, along with all earlier public Swift data since 2005.
During this time interval, OJ 287 was caught in extreme minima and outburst
states. Its X-ray spectrum is highly variable and encompasses all states seen
in blazars from very flat to exceptionally steep. The spectrum can be
decomposed into three spectral components: Inverse Compton (IC) emission
dominant at low-states, super-soft synchrotron emission which becomes
increasingly dominant as OJ 287 brightens, and an intermediately-soft
(Gamma_x=2.2) additional component seen at outburst. This last component
extends beyond 10 keV and plausibly represents either a second synchrotron/IC
component and/or a temporary disk corona of the primary supermassive black hole
(SMBH). Our 2018 XMM-Newton observation, quasi-simultaneous with the Event
Horizon Telescope observation of OJ 287, is well described by a two-component
model with a hard IC component of Gamma_x=1.5 and a soft synchrotron component.
Low-state spectra limit any long-lived accretion disk/corona contribution in
X-rays to a very low value of L_x/L_Edd < 5.6 times 10^(-4) (for M_(BH,
primary) = 1.8 times 10^10 M_sun). Some implications for the binary SMBH model
of OJ 287 are discussed.
|
Several recent applications of optimal transport (OT) theory to machine
learning have relied on regularization, notably entropy and the Sinkhorn
algorithm. Because matrix-vector products are pervasive in the Sinkhorn
algorithm, several works have proposed to \textit{approximate} kernel matrices
appearing in its iterations using low-rank factors. Another route lies instead
in imposing low-rank constraints on the feasible set of couplings considered in
OT problems, with no approximations on cost nor kernel matrices. This route was
first explored by Forrow et al., 2018, who proposed an algorithm tailored for
the squared Euclidean ground cost, using a proxy objective that can be solved
through the machinery of regularized 2-Wasserstein barycenters. Building on
this, we introduce in this work a generic approach that aims at solving, in
full generality, the OT problem under low-rank constraints with arbitrary
costs. Our algorithm relies on an explicit factorization of low rank couplings
as a product of \textit{sub-coupling} factors linked by a common marginal;
similar to an NMF approach, we alternatively updates these factors. We prove
the non-asymptotic stationary convergence of this algorithm and illustrate its
efficiency on benchmark experiments.
|
The hierarchy of nonlocality and entanglement in multipartite systems is one
of the fundamental problems in quantum physics. Existing studies on this topic
to date were limited to the entanglement classification according to the
numbers of particles enrolled. Equivalence under stochastic local operations
and classical communication provides a more detailed classification, e. g. the
genuine three-qubit entanglement being divided into W and GHZ classes. We
construct two families of local models for the three-qubit
Greenberger-Horne-Zeilinger (GHZ)-symmetric states, whose entanglement classes
have a complete description. The key technology of construction the local
models in this work is the GHZ symmetrization on tripartite extensions of the
optimal local-hidden-state models for Bell diagonal states. Our models show
that entanglement and nonlocality are inequivalent for all the entanglement
classes (biseparable, W, and GHZ) in three-qubit systems.
|
We show that both the classical as well as the quantum definitions of the
Fisher information faithfully identify resourceful quantum states in general
quantum resource theories, in the sense that they can always distinguish
between states with and without a given resource. This shows that all quantum
resources confer an advantage in metrology, and establishes the Fisher
information as a universal tool to probe the resourcefulness of quantum states.
We provide bounds on the extent of this advantage, as well as a simple
criterion to test whether different resources are useful for the estimation of
unitarily encoded parameters. Finally, we extend the results to show that the
Fisher information is also able to identify the dynamical resourcefulness of
quantum operations.
|
The relation of period spacing ($\Delta P$) versus period ($P$) of dipole
prograde g modes is known to be useful to measure rotation rates in the g-mode
cavity of rapidly rotating $\gamma$ Dor and slowly pulsating B (SPB) stars. In
a rapidly rotating star, an inertial mode in the convective core can resonantly
couple with g modes propagative in the surrounding radiative region. The
resonant coupling causes a dip in the $P$-$\Delta P$ relation, distinct from
the modulations due to the chemical composition gradient. Such a resonance dip
in $\Delta P$ of prograde dipole g modes appears around a frequency
corresponding to a spin parameter $2f_{\rm rot}{\rm(cc)}/\nu_{\rm co-rot} \sim
8-11$ with $f_{\rm rot}$(cc) being the rotation frequency of the convective
core and $\nu_{\rm co-rot}$ the pulsation frequency in the co-rotating frame.
The spin parameter at the resonance depends somewhat on the extent of core
overshooting, central hydrogen abundance, and other stellar parameters. We can
fit the period at the observed dip with the prediction from prograde dipole g
modes of a main-sequence model, allowing the convective core to rotate
differentially from the surrounding g-mode cavity. We have performed such
fittings for 16 selected $\gamma$ Dor stars having well defined dips, and found
that the majority of $\gamma$ Dor stars we studied rotate nearly uniformly,
while convective cores tend to rotate slightly faster than the g-mode cavity in
less evolved stars.
|
Magnetic reconnection is explored on the Terrestrial Reconnection Experiment
(TREX) for asymmetric inflow conditions and in a configuration where the
absolute rate of reconnection is set by an external drive. Magnetic pileup
enhances the upstream magnetic field of the high density inflow, leading to an
increased upstream Alfven speed and helping to lower the normalized
reconnection rate to values expected from theoretical consideration. In
addition, a shock interface between the far upstream supersonic plasma inflow
and the region of magnetic flux pileup is observed, important to the overall
force balance of the system, hereby demonstrating the role of shock formation
for configurations including a supersonically driven inflow. Despite the
specialised geometry where a strong reconnection drive is applied from only one
side of the reconnection layer, previous numerical and theoretical results
remain robust and are shown to accurately predict the normalized rate of
reconnection for the range of system sizes considered. This experimental rate
of reconnection is dependent on system size, reaching values as high as 0.8 at
the smallest normalized system size applied.
|
In the de Sitter gauge theory (DGT), the fundamental variables are the de
Sitter (dS) connection and the gravitational Higgs/Goldstone field $\xi^A$.
Previously, a model for DGT was analyzed, which generalizes the
MacDowell--Mansouri gravity to have a variable cosmological constant
$\Lambda=3/l^2$, where $l$ is related to $\xi^A$ by $\xi^A\xi_A=l^2$. It was
shown that the model sourced by a perfect fluid does not support a radiation
epoch and the accelerated expansion of the parity invariant universe. In this
work, I consider a similar model, namely, the Stelle--West gravity, and couple
it to a modified perfect fluid, such that the total Lagrangian 4-form is
polynomial in the gravitational variables. The Lagrangian of the modified fluid
has a nontrivial variational derivative with respect to $l$, and as a result,
the problems encountered in the previous work no longer appear. Moreover, to
explore the elegance of the general theory, as well as to write down the basic
framework, I perform the Lagrange--Noether analysis for DGT sourced by a matter
field, yielding the field equations and the identities with respect to the
symmetries of the system. The resulted formula are dS covariant and do not rely
on the existence of the metric field.
|
Privacy and energy are primary concerns for sensor devices that offload
compute to a potentially untrusted edge server or cloud. Homomorphic Encryption
(HE) enables offload processing of encrypted data. HE offload processing
retains data privacy, but is limited by the need for frequent communication
between the client device and the offload server. Existing client-aided
encrypted computing systems are optimized for performance on the offload
server, failing to sufficiently address client costs, and precluding HE offload
for low-resource (e.g., IoT) devices. We introduce Client-aided HE for Opaque
Compute Offloading (CHOCO), a client-optimized system for encrypted offload
processing. CHOCO introduces rotational redundancy, an algorithmic optimization
to minimize computing and communication costs. We design Client-Aided HE for
Opaque Compute Offloading Through Accelerated Cryptographic Operations
(CHOCO-TACO), a comprehensive architectural accelerator for client-side
cryptographic operations that eliminates most of their time and energy costs.
Our evaluation shows that CHOCO makes client-aided HE offloading feasible for
resource-constrained clients. Compared to existing encrypted computing
solutions, CHOCO reduces communication cost by up to 2948x. With hardware
support, client-side encryption/decryption is faster by 1094x and uses 648x
less energy. In our end-to-end implementation of a large-scale DNN (VGG16),
CHOCO uses 37% less energy than local (unencrypted) computation.
|
Primordial perturbations in our universe are believed to have a quantum
origin, and can be described by the wavefunction of the universe (or
equivalently, cosmological correlators). It follows that these observables must
carry the imprint of the founding principle of quantum mechanics: unitary time
evolution. Indeed, it was recently discovered that unitarity implies an
infinite set of relations among tree-level wavefunction coefficients, dubbed
the Cosmological Optical Theorem. Here, we show that unitarity leads to a
systematic set of "Cosmological Cutting Rules" which constrain wavefunction
coefficients for any number of fields and to any loop order. These rules fix
the discontinuity of an n-loop diagram in terms of lower-loop diagrams and the
discontinuity of tree-level diagrams in terms of tree-level diagrams with fewer
external fields. Our results apply with remarkable generality, namely for
arbitrary interactions of fields of any mass and any spin with a Bunch-Davies
vacuum around a very general class of FLRW spacetimes. As an application, we
show how one-loop corrections in the Effective Field Theory of inflation are
fixed by tree-level calculations and discuss related perturbative unitarity
bounds. These findings greatly extend the potential of using unitarity to
bootstrap cosmological observables and to restrict the space of consistent
effective field theories on curved spacetimes.
|
In this paper, we show how the absorption and re-radiation energy from
molecules in the air can influence the Multiple Input Multiple Output (MIMO)
performance in high-frequency bands, e.g., millimeter wave (mmWave) and
terahertz. In more detail, some common atmosphere molecules, such as oxygen and
water, can absorb and re-radiate energy in their natural resonance frequencies,
such as 60 GHz, 180 GHz and 320 GHz. Hence, when hit by electromagnetic waves,
molecules will get excited and absorb energy, which leads to an extra path loss
and is known as molecular attenuation. Meanwhile, the absorbed energy will be
re-radiated towards a random direction with a random phase. These re-radiated
waves also interfere with the signal transmission. Although, the molecular
re-radiation was mostly considered as noise in literature, recent works show
that it is correlated to the main signal and can be viewed as a composition of
multiple delayed or scattered signals. Such a phenomenon can provide
non-line-of-sight (NLoS) paths in an environment that lacks scatterers, which
increases spatial multiplexing and thus greatly enhances the performance of
MIMO systems. Therefore in this paper, we explore the scattering model and
noise models of molecular re-radiation to characterize the channel transfer
function of the NLoS channels created by atmosphere molecules. Our simulation
results show that the re-radiation can increase MIMO capacity up to 3 folds in
mmWave and 6 folds in terahertz for a set of realistic transmit power,
distance, and antenna numbers. We also show that in the high SNR, the
re-radiation makes the open-loop precoding viable, which is an alternative to
beamforming to avoid beam alignment sensitivity in high mobility applications.
|
We present a Hubble Space Telescope/Wide-Field Camera 3 near infrared
spectrum of the archetype Y dwarf WISEP 182831.08+265037.8. The spectrum covers
the 0.9-1.7 um wavelength range at a resolving power of lambda/Delta lambda
~180 and is a significant improvement over the previously published spectrum
because it covers a broader wavelength range and is uncontaminated by light
from a background star. The spectrum is unique for a cool brown dwarf in that
the flux peaks in the Y, J, and H band are of near equal intensity in units of
f_lambda. We fail to detect any absorption bands of NH_3 in the spectrum, in
contrast to the predictions of chemical equilibrium models, but tentatively
identify CH_4 as the carrier of an unknown absorption feature centered at 1.015
um. Using previously published ground- and spaced-based photometry, and using a
Rayleigh Jeans tail to account for flux emerging longward of 4.5 um, we compute
a bolometric luminosity of log (L_bol/L_sun)=-6.50+-0.02 which is significantly
lower than previously published results. Finally, we compare the spectrum and
photometry to two sets of atmospheric models and find that best overall match
to the observed properties of WISEP 182831.08+265037.8 is a ~1 Gyr old binary
composed of two T_eff~325 K, ~5 M_Jup brown dwarfs with subsolar [C/O] ratios.
|
Consider traffic data (i.e., triplets in the form of
source-destination-timestamp) that grow over time. Tensors (i.e.,
multi-dimensional arrays) with a time mode are widely used for modeling and
analyzing such multi-aspect data streams. In such tensors, however, new entries
are added only once per period, which is often an hour, a day, or even a year.
This discreteness of tensors has limited their usage for real-time
applications, where new data should be analyzed instantly as it arrives. How
can we analyze time-evolving multi-aspect sparse data 'continuously' using
tensors where time is'discrete'? We propose SLICENSTITCH for continuous
CANDECOMP/PARAFAC (CP) decomposition, which has numerous time-critical
applications, including anomaly detection, recommender systems, and stock
market prediction. SLICENSTITCH changes the starting point of each period
adaptively, based on the current time, and updates factor matrices (i.e.,
outputs of CP decomposition) instantly as new data arrives. We show,
theoretically and experimentally, that SLICENSTITCH is (1) 'Any time': updating
factor matrices immediately without having to wait until the current time
period ends, (2) Fast: with constant-time updates up to 464x faster than online
methods, and (3) Accurate: with fitness comparable (specifically, 72 ~ 100%) to
offline methods.
|
We develop a first-principles-based generalized mode-coupling theory (GMCT)
for the tagged-particle motion of glassy systems. This theory establishes a
hierarchy of coupled integro-differential equations for self-multi-point
density correlation functions, which can formally be extended up to infinite
order. We use our GMCT framework to calculate the self-nonergodicity parameters
and the self-intermediate scattering function for the Percus-Yevick hard sphere
system, based on the first few levels of the GMCT hierarchy. We also test the
scaling laws in the $\alpha$- and $\beta$-relaxation regimes near the
glass-transition singularity. Furthermore, we study the mean-square
displacement and the Stoke-Einstein relation in the supercooled regime. We find
that qualitatively our GMCT results share many similarities with the
well-established predictions from standard mode-coupling theory, but the
quantitative results change, and typically improve, by increasing the GMCT
closure level. However, we also demonstrate on general theoretical grounds that
the current GMCT framework is unable to account for violation of the
Stokes-Einstein relation, underlining the need for further improvements in the
first-principles description of glassy dynamics.
|
In this article, we define amorphic complexity for actions of locally compact
$\sigma$-compact amenable groups on compact metric spaces. Amorphic complexity,
originally introduced for $\mathbb Z$-actions, is a topological invariant which
measures the complexity of dynamical systems in the regime of zero entropy. We
show that it is tailor-made to study strictly ergodic group actions with
discrete spectrum and continuous eigenfunctions. This class of actions
includes, in particular, Delone dynamical systems related to regular model sets
obtained via Meyer's cut and project method. We provide sharp upper bounds on
amorphic complexity of such systems. In doing so, we observe an intimate
relationship between amorphic complexity and fractal geometry.
|
In this paper we prove regularity results for a class of nonlinear degenerate
elliptic equations of the form $\displaystyle -\operatorname{div}(A(|\nabla
u|)\nabla u)+B\left( |\nabla u|\right) =f(u)$; in particular, we investigate
the second order regularity of the solutions. As a consequence of these
results, we obtain symmetry and monotonicity properties of positive solutions
for this class of degenerate problems in convex symmetric domains via a
suitable adaption of the celebrated moving plane method of Alexandrov-Serrin.
|
In the centre of mass frame, we have studied theoretically the $Z$-boson
resonant production in the presence of an intense laser field via the weak
process $e^+e^- \to \mu^+\mu^-$. Dressing the incident particles by a
Circularly Polarized laser field (CP-laser field), at the first step, shows
that for a given laser field's parameters, the $Z$- boson cross section
decreases by several orders of magnitude. We have compared the the Total Cross
Section (TCS) obtained by using the scattering matrix method with that given by
the Breit-Wigner approach in the presence of a CP-laser field and the results
are found to be very consistent. This result indicates that Breit-Wigner
formula is valid not only for the laser-free process but also in the presence
of a CP-laser field. The dependence of the laser-assisted differential cross
section on the Centre of Mass Energy (CME) for different scattering angles
proves that it reaches its maximum for small and high scattering angles. At the
next step and by dressing both incident and scattered particles, we have shown
that the CP-laser field largely affects the TCS, especially when its strength
reaches $10^{9}\,V.cm^{-1}$. This result confirms that obtained for the elastic
electron-proton scattering in the presence of a CP-laser field [I. Dahiri et
al., arXiv:2102.00722v1]. It is interpreted by the fact that heavy interacting
particles require high laser field's intensity to affect the collision's cross
section.
|
In this work, we give a new technique for constructing self-dual codes over
commutative Frobenius rings using $\lambda$-circulant matrices. The new
construction was derived as a modification of the well-known four circulant
construction of self-dual codes. Applying this technique together with the
building-up construction, we construct singly-even binary self-dual codes of
lengths 56, 58, 64, 80 and 92 that were not known in the literature before.
Singly-even self-dual codes of length 80 with $\beta\in\{2,4,5,6,8\}$ in their
weight enumerators are constructed for the first time in the literature.
|
The need for a comprehensive study to explore various aspects of online
social media has been instigated by many researchers. This paper gives an
insight into the social platform, Twitter. In this present work, we have
illustrated stepwise procedure for crawling the data and discuss the key issues
related to extracting associated features that can be useful in Twitter-related
research while crawling these data from Application Programming Interfaces
(APIs). Further, the data that comprises of over 86 million tweets have been
analysed from various perspective including the most used languages, most
frequent words, most frequent users, countries with most and least tweets and
re-tweets, etc. The analysis reveals that the users' data associated with
Twitter has a high affinity for researches in the various domain that includes
politics, social science, economics, and linguistics, etc. In addition, the
relation between Twitter users of a country and its human development index has
been identified. It is observed that countries with very high human development
indices have a relatively higher number of tweets compared to low human
development indices countries. It is envisaged that the present study shall
open many doors of researches in information processing and data science.
|
We study the structural evolution of isolated star-forming galaxies in the
Illustris TNG100-1 hydrodynamical simulation, with a focus on investigating the
growth of the central core density within 2 kpc ($\Sigma_{*,2kpc}$) in relation
to total stellar mass ($M_*$) at z < 0.5. First, we show that several
observational trends in the $\Sigma_{*,2kpc}$-$M_*$ plane are qualitatively
reproduced in IllustrisTNG, including the distributions of AGN, star forming
galaxies, quiescent galaxies, and radial profiles of stellar age, sSFR, and
metallicity. We find that galaxies with dense cores evolve parallel to the
$\Sigma_{*,2kpc}$-$M_*$ relation, while galaxies with diffuse cores evolve
along shallower trajectories. We investigate possible drivers of rapid growth
in $\Sigma_{*,2kpc}$ compared to $M_*$. Both the current sSFR gradient and the
BH accretion rate are indicators of past core growth, but are not predictors of
future core growth. Major mergers (although rare in our sample; $\sim$10%)
cause steeper core growth, except for high mass ($M_*$ >$\sim$ $10^{10}
M_{\odot}$) mergers, which are mostly dry. Disc instabilities, as measured by
the fraction of mass with Toomre Q < 2, are not predictive of rapid core
growth. Instead, rapid core growth results in more stable discs. The cumulative
black hole feedback history sets the maximum rate of core growth, preventing
rapid growth in high-mass galaxies ($M_*$ >$\sim$ $10^{9.5} M_{\odot}$). For
massive galaxies the total specific angular momentum of accreting gas is the
most important predictor of future core growth. Our results suggest that the
angular momentum of accreting gas controls the slope, width and zero-point
evolution of the $\Sigma_{*,2kpc}$-$M_*$ relation.
|
This article introduces a neural network-based signal processing framework
for intelligent reflecting surface (IRS) aided wireless communications systems.
By modeling radio-frequency (RF) impairments inside the "meta-atoms" of IRS
(including nonlinearity and memory effects), we present an approach that
generalizes the entire IRS-aided system as a reservoir computing (RC) system,
an efficient recurrent neural network (RNN) operating in a state near the "edge
of chaos". This framework enables us to take advantage of the nonlinearity of
this "fabricated" wireless environment to overcome link degradation due to
model mismatch. Accordingly, the randomness of the wireless channel and RF
imperfections are naturally embedded into the RC framework, enabling the
internal RC dynamics lying on the edge of chaos. Furthermore, several practical
issues, such as channel state information acquisition, passive beamforming
design, and physical layer reference signal design, are discussed.
|
Is critical input information encoded in specific sparse pathways within the
neural network? In this work, we discuss the problem of identifying these
critical pathways and subsequently leverage them for interpreting the network's
response to an input. The pruning objective -- selecting the smallest group of
neurons for which the response remains equivalent to the original network --
has been previously proposed for identifying critical pathways. We demonstrate
that sparse pathways derived from pruning do not necessarily encode critical
input information. To ensure sparse pathways include critical fragments of the
encoded input information, we propose pathway selection via neurons'
contribution to the response. We proceed to explain how critical pathways can
reveal critical input features. We prove that pathways selected via neuron
contribution are locally linear (in an L2-ball), a property that we use for
proposing a feature attribution method: "pathway gradient". We validate our
interpretation method using mainstream evaluation experiments. The validation
of pathway gradient interpretation method further confirms that selected
pathways using neuron contributions correspond to critical input features. The
code is publicly available.
|
The paper investigates the problem of finding communities in complex network
systems, the detection of which allows a better understanding of the laws of
their functioning. To solve this problem, two approaches are proposed based on
the use of flows characteristics of complex network. The first of these
approaches consists in calculating the parameters of influence of separate
subsystems of the network system, distinguished by the principles of ordering
or subordination, and the second, in using the concept of its flow core. Based
on the proposed approaches, reliable criteria for finding communities have been
formulated and efficient algorithms for their detection in complex network
systems have been developed. It is shown that the proposed approaches make it
possible to single out communities in cases in which the existing numerical and
visual methods turn out to be disabled.
|
The novel concept of simultaneously transmitting and reflecting (STAR)
reconfigurable intelligent surfaces (RISs) is investigated, where the incident
wireless signal is divided into transmitted and reflected signals passing into
both sides of the space surrounding the surface, thus facilitating a full-space
manipulation of signal propagation. Based on the introduced basic signal model
of `STAR', three practical operating protocols for STAR-RISs are proposed,
namely energy splitting (ES), mode switching (MS), and time switching (TS).
Moreover, a STAR-RIS aided downlink communication system is considered for both
unicast and multicast transmission, where a multi-antenna base station (BS)
sends information to two users, i.e., one on each side of the STAR-RIS. A power
consumption minimization problem for the joint optimization of the active
beamforming at the BS and the passive transmission and reflection beamforming
at the STAR-RIS is formulated for each of the proposed operating protocols,
subject to communication rate constraints of the users. For ES, the resulting
highly-coupled non-convex optimization problem is solved by an iterative
algorithm, which exploits the penalty method and successive convex
approximation. Then, the proposed penalty-based iterative algorithm is extended
to solve the mixed-integer non-convex optimization problem for MS. For TS, the
optimization problem is decomposed into two subproblems, which can be
consecutively solved using state-of-the-art algorithms and convex optimization
techniques. Finally, our numerical results reveal that: 1) the TS and ES
operating protocols are generally preferable for unicast and multicast
transmission, respectively; and 2) the required power consumption for both
scenarios is significantly reduced by employing the proposed STAR-RIS instead
of conventional reflecting/transmiting-only RISs.
|
In this work we compute the first integral cohomology of the pure mapping
class group of a non-orientable surface of infinite topological type and genus
at least 3. To this purpose, we also prove several other results already known
for orientable surfaces such as the existence of an Alexander method, the fact
that the mapping class group is isomorphic to the automorphism group of the
curve graph along with the topological rigidity of the curve graph, and the
structure of the pure mapping class group as both a Polish group and a
semi-direct product.
|
Programmable data planes allow users to define their own data plane
algorithms for network devices including appropriate data plane application
programming interfaces (APIs) which may be leveraged by user-defined
software-defined networking (SDN) control. This offers great flexibility for
network customization, be it for specialized, commercial appliances, e.g., in
5G or data center networks, or for rapid prototyping in industrial and academic
research. Programming protocol-independent packet processors (P4) has emerged
as the currently most widespread abstraction, programming language, and concept
for data plane programming. It is developed and standardized by an open
community, and it is supported by various software and hardware platforms. In
the first part of this paper we give a tutorial of data plane programming
models, the P4 programming language, architectures, compilers, targets, and
data plane APIs. We also consider research efforts to advance P4 technology. In
the second part, we categorize a large body of literature of P4-based applied
research into different research domains, summarize the contributions of these
papers, and extract prototypes, target platforms, and source code availability.
For each research domain, we analyze how the reviewed works benefit from P4's
core features. Finally, we discuss potential next steps based on our findings.
|
Metasurfaces enable manipulation of light propagation at an unprecedented
level, benefitting from a number of merits unavailable to conventional optical
elements, such as ultracompactness, precise phase and polarization control at
deep subwavelength scale, and multifunctionalities. Recent progress in this
field has witnessed a plethora of functional metasurfaces, ranging from lenses
and vortex beam generation to holography. However, research endeavors have been
mainly devoted to static devices, exploiting only a glimpse of opportunities
that metasurfaces can offer. We demonstrate a dynamic metasurface platform,
which allows independent manipulation of addressable subwavelength pixels at
visible frequencies through controlled chemical reactions. In particular, we
create dynamic metasurface holograms for advanced optical information
processing and encryption. Plasmonic nanorods tailored to exhibit hierarchical
reaction kinetics upon hydrogenation/dehydrogenation constitute addressable
pixels in multiplexed metasurfaces. The helicity of light, hydrogen, oxygen,
and reaction duration serve as multiple keys to encrypt the metasurfaces. One
single metasurface can be deciphered into manifold messages with customized
keys, featuring a compact data storage scheme as well as a high level of
information security. Our work suggests a novel route to protect and transmit
classified data, where highly restricted access of information is imposed.
|
High-energy heavy-ion collisions generate extremely strong magnetic field
which plays a key role in a number of novel quantum phenomena in quark-gluon
plasma (QGP), such as the chiral magnetic effect (CME). However, due to the
complexity in theoretical modellings of the coupled electromagnetic fields and
the QGP system, especially in the pre-equilibrium stages, the lifetime of the
magnetic field in the QGP medium remains undetermined. We establish, for the
first time, a kinetic framework to study the dynamical decay of the magnetic
field in the early stages of a weakly coupled QGP by solving the coupled
Boltzmann and Maxwell equations. We find that at late times a
magnetohydrodynamical description of the coupled system emerges. With respect
to realistic collisions at RHIC and the LHC, we estimate the residual strength
of the magnetic field in the QGP when the system start to evolve
hydrodynamically.
|
Nowadays, the confidentiality of data and information is of great importance
for many companies and organizations. For this reason, they may prefer not to
release exact data, but instead to grant researchers access to approximate
data. For example, rather than providing the exact measurements of their
clients, they may only provide researchers with grouped data, that is, the
number of clients falling in each of a set of non-overlapping measurement
intervals. The challenge is to estimate the mean and variance structure of the
hidden ungrouped data based on the observed grouped data. To tackle this
problem, this work considers the exact observed data likelihood and applies the
Expectation-Maximization (EM) and Monte-Carlo EM (MCEM) algorithms for cases
where the hidden data follow a univariate, bivariate, or multivariate normal
distribution. Simulation studies are conducted to evaluate the performance of
the proposed EM and MCEM algorithms. The well-known Galton data set is
considered as an application example.
|
For the first time, the dielectric response of a BaTiO3 thin film under an AC
electric field is investigated using time-resolved X-ray absorption
spectroscopy at the Ti K-edge to clarify correlated contributions of each
constituent atom on the electronic states. Intensities of the pre-edge eg peak
and shoulder structure just below the main edge increase with an increase in
the amplitude of the applied electric field, whereas that of the main peak
decreases in an opposite manner. Based on the multiple scattering theory, the
increase and decrease of the eg and main peaks are simulated for different Ti
off-center displacements. Our results indicate that these spectral features
reflect the inter- and intra-atomic hybridization of Ti 3d with O 2p and Ti 4p,
respectively. In contrast, the shoulder structure is not affected by changes in
the Ti off-center displacement but is susceptible to the effect of the corner
site Ba ions. This is the first experimental verification of the dynamic
electronic contribution of Ba to polarization reversal.
|
Magnetic multilayers are promising tuneable systems for hosting magnetic
skyrmions at/above room temperature. Revealing their intriguing switching
mechanisms and associated inherent electrical responses are prerequisites for
developing skyrmionic devices. In this work, we theoretically demonstrate the
annihilation of single skyrmions occurring through a multilayer structure,
which is mediated by hopping dynamics of topological hedgehog singularities
known as Bloch points. The emerging intralayer dynamics of Bloch points are
dominated by the Dzyaloshinskii-Moriya interaction, and their propagation can
give rise to solenoidal emergent electric fields in the vicinity. Moreover, as
the topology of spin textures can dominate their emergent magnetic properties,
we show that the Bloch-point hopping through the multilayer will modulate the
associated topological Hall response, with the magnitude proportional to the
effective topological charge. We also investigate the thermodynamic stability
of these states regarding the layer-dependent magnetic properties. This study
casts light on the emergent electromagnetic signatures of skyrmion-based
spintronics, rooted in magnetic-multilayer systems.
|
In this paper a comparative structural, dielectric and magnetic study of two
langasite compounds Ba$_3$TeCo$_3$P$_2$O$_{14}$ (absence of lone pair) and
Pb$_3$TeCo$_3$P$_2$O$_{14}$ (Pb$^{2+}$ 6$s^2$ lone pair) have been carried out
to precisely explore the development of room temperature spontaneous
polarization in presence of stereochemically active lone pair. In case of
Pb$_3$TeCo$_3$P$_2$O$_{14}$, mixing of both Pb 6$s$ with Pb 6$p$ and O 2$p$
help the lone pair to be stereochemically active. This stereochemically active
lone pair brings a large structural distortion within the unit cell and creates
a polar geometry, while Ba$_3$TeCo$_3$P$_2$O$_{14}$ compound remains in a
nonpolar structure due to the absence of any such effect. Consequently,
polarization measurement under varying electric field confirms room temperature
ferroelectricity for Pb$_3$TeCo$_3$P$_2$O$_{14}$, which was not the case of
Ba$_3$TeCo$_3$P$_2$O$_{14}$. Detailed study was carried out to understand the
microscopic mechanism of ferroelectricity which revealed the exciting
underlying activity of poler TeO$_6$ octahedral unit as well as Pb-hexagon.
|
We study cosmological inflation and its dynamics in the framework of the
Randall-Sundrum II brane model. In particular, we analyze in detail four
representative small-field inflationary potentials, namely Natural inflation,
Hilltop inflation, Higgs-like inflation, and Exponential SUSY inflation, each
characterized by two mass scales. We constrain the parameters for which a
viable inflationary Universe emerges using the latest PLANCK results.
Furthermore, we investigate whether or not those models in brane cosmology are
consistent with the recently proposed Swampland Criteria, and give predictions
for the duration of reheating as well as for the reheating temperature after
inflation. Our results show that (i) the distance conjecture is satisfied, (ii)
the de Sitter conjecture and its refined version may be avoided, and (iii) the
allowed range for the five-dimensional Planck mass, $M_5$, is found to be
between $10^5~\textrm{TeV}$ and $10^{12}~\textrm{TeV}$. Our main findings
indicate that non-thermal leptogenesis cannot work within the framework of
RS-II brane cosmology, at least for the inflationary potentials considered
here.
|
We first propose a general method to construct the complete set of on-shell
operator bases involving massive particles with any spins. To incorporate the
non-abelian little groups of massive particles, the on-shell scattering
amplitude basis should be factorized into two parts: one is charged, and the
other one is neutral under little groups of massive particles. The complete set
of these two parts can be systematically constructed by choosing some specific
Young diagrams of Lorentz subgroup and global symmetry $U(N)$ respectively ($N$
is the number of external particles), without the equation of motion and
integration by part redundancy. Thus the complete massive amplitude bases
without any redundancies can be obtained by combining these two complete sets.
Some examples are presented to explicitly demonstrate this method. This method
is applicable for constructing amplitude bases involving identical particles,
and all the bases can be constructed automatically by computer programs based
on it.
|
We study the variety ZG of monoids where the elements that belong to a group
are central, i.e., commute with all other elements. We show that ZG is local,
that is, the semidirect product ZG * D of ZG by definite semigroups is equal to
LZG, the variety of semigroups where all local monoids are in ZG. Our main
result is thus: ZG * D = LZG. We prove this result using Straubing's delay
theorem, by considering paths in the category of idempotents. In the process,
we obtain the characterization ZG = MNil \vee Com, and also characterize the ZG
languages, i.e., the languages whose syntactic monoid is in ZG: they are
precisely the languages that are finite unions of disjoint shuffles of
singleton languages and regular commutative languages.
|
The noise-enhanced trapping is a surprising phenomenon that has already been
studied in chaotic scattering problems where the noise affects the physical
variables but not the parameters of the system. Following this research, in
this work we provide strong numerical evidence to show that an additional
mechanism that enhances the trapping arises when the noise influences the
energy of the system. For this purpose, we have included a source of Gaussian
white noise in the H\'enon-Heiles system, which is a paradigmatic example of
open Hamiltonian system. For a particular value of the noise intensity, some
trajectories decrease their energy due to the stochastic fluctuations. This
drop in energy allows the particles to spend very long transients in the
scattering region, increasing their average escape times. This result, together
with the previously studied mechanisms, points out the generality of the
noise-enhanced trapping in chaotic scattering problems.
|
This study investigates the correlation of self-report accuracy with academic
performance. The sample was composed of 289 undergraduate students (96 senior
and 193 junior) enrolled in two engineering classes. Age ranged between 22 and
24 years, with a slight over representation of male students (53%). Academic
performance was calculated based on students' final grades in each class. The
tendency to report inaccurate information was measured at the end of the Raven
Progressive Matrices Test, by asking students to report their exact finishing
times. We controlled for gender, age, personality traits, intelligence, and
past academic performance. We also included measures of centrality in their
friendship, advice and trust networks. Correlation and multiple regression
analyses results indicate that lower achieving students were significantly less
accurate in self-reporting data. We also found that being more central in the
advice network was correlated with higher performance (r = .20, p < .001). The
results are aligned with existing literature emphasizing the individual and
relational factors associated with academic performance and, pending future
studies, may be utilized to include a new metric of self-report accuracy that
is not dependent on academic records.
|
We investigate the 3D spin alignment of galaxies with respect to the
large-scale filaments using the MaNGA survey. The cosmic web is reconstructed
from the Sloan Digital Sky Survey using Disperse and the 3D spins of MaNGA
galaxies are estimated using the thin disk approximation with integral field
spectroscopy kinematics. Late-type spiral galaxies are found to have their
spins parallel to the closest filament's axis. The alignment signal is found to
be dominated by low-mass spirals. Spins of S0-type galaxies tend to be oriented
preferentially in perpendicular direction with respect to the filament's axis.
This orthogonal orientation is found to be dominated by S0s that show a notable
misalignment between their kinematic components of stellar and ionised gas
velocity fields and/or by low mass S0s with lower rotation support compared to
their high mass counterparts. Qualitatively similar results are obtained when
splitting galaxies based on the degree of ordered stellar rotation, such that
galaxies with high spin magnitude have their spin aligned, and those with low
spin magnitude in perpendicular direction to the filaments. In the context of
conditional tidal torque theory, these findings suggest that galaxies' spins
retain memory of their larger-scale environment. In agreement with measurements
from hydrodynamical cosmological simulations, the measured signal at low
redshift is weak, yet statistically significant. The dependence of the
spin-filament orientation of galaxies on their stellar mass, morphology and
kinematics highlights the importance of sample selection to detect the signal.
|
Pretrained Masked Language Models (MLMs) have revolutionised NLP in recent
years. However, previous work has indicated that off-the-shelf MLMs are not
effective as universal lexical or sentence encoders without further
task-specific fine-tuning on NLI, sentence similarity, or paraphrasing tasks
using annotated task data. In this work, we demonstrate that it is possible to
turn MLMs into effective universal lexical and sentence encoders even without
any additional data and without any supervision. We propose an extremely
simple, fast and effective contrastive learning technique, termed Mirror-BERT,
which converts MLMs (e.g., BERT and RoBERTa) into such encoders in 20-30
seconds without any additional external knowledge. Mirror-BERT relies on fully
identical or slightly modified string pairs as positive (i.e., synonymous)
fine-tuning examples, and aims to maximise their similarity during identity
fine-tuning. We report huge gains over off-the-shelf MLMs with Mirror-BERT in
both lexical-level and sentence-level tasks, across different domains and
different languages. Notably, in the standard sentence semantic similarity
(STS) tasks, our self-supervised Mirror-BERT model even matches the performance
of the task-tuned Sentence-BERT models from prior work. Finally, we delve
deeper into the inner workings of MLMs, and suggest some evidence on why this
simple approach can yield effective universal lexical and sentence encoders.
|
Soft or weakly-consolidated sand refers to porous materials composed of
particles (or grains) weakly held together to form a solid but that can be
easily broken when subjected to stress. These materials do not behave as
conventional brittle, linear elastic materials and the transition between these
two regimes cannot usually be described using poro-elastic models. Furthermore,
conventional geotechnical sampling techniques often result in the destruction
of the cementation and recovery of sufficient intact core is, therefore,
difficult. This paper studies a numerical model that allows us to introduce
weak consolidation in granular packs. The model, based on the LIGGGHTS open
source project, simply adds an attractive contribution to particles in contact.
This simple model allow us to reproduce key elements of the behaviour of the
stress observed in compacted sands and clay, as well as in poorly consolidated
sandstones. The paper finishes by inspecting the effect of different
consolidation levels in fluid-driven fracture behaviour. Numerical results are
compared against experimental results on bio-cemented sandstones.
|
In this paper, a new implicit-explicit local method with an arbitrary order
is produced for stiff initial value problems. Here, a general method for
one-step time integrations has been created, considering a direction free
approach for integrations leading to a numerical method with parameter-based
stability preservation. Adaptive procedures depending on the problem types for
the current method are explained with the help of local error estimates to
minimize the computational cost. Priority error analysis of the current method
is made, and order conditions are presented in terms of direction parameters.
Stability analysis of the method is performed for both scalar equations and
systems of differential equations. The currently produced parameter-based
method has been proven to provide A-stability, for 0.5<\theta<1, in various
orders. The present method has been shown to be a very good option for
addressing a wide range of initial value problems through numerical
experiments. It can be seen as a significant contribution that the
Susceptible-Exposed-Infected-Recovered equation system parameterized for the
COVID-19 pandemic has been integrated with the present method and stability
properties of the method have been tested on this stiff model and significant
results are produced. Some challenging stiff behaviours represented by the
nonlinear Duffing equation, Robertson chemical system, and van der Pol equation
have also been integrated, and the results revealed that the current algorithm
produces much more reliable results than numerical techniques in the
literature.
|
In this paper, a new method of training pipeline is discussed to achieve
significant performance on the task of anti-spoofing with RGB image. We explore
and highlight the impact of using pseudo-depth to pre-train a network that will
be used as the backbone to the final classifier. While the usage of
pseudo-depth for anti-spoofing task is not a new idea on its own, previous
endeavours utilize pseudo-depth simply as another medium to extract features
for performing prediction, or as part of many auxiliary losses in aiding the
training of the main classifier, normalizing the importance of pseudo-depth as
just another semantic information. Through this work, we argue that there
exists a significant advantage in training the final classifier can be gained
by the pre-trained generator learning to predict the corresponding pseudo-depth
of a given facial image, from a Generative Adversarial Network framework. Our
experimental results indicate that our method results in a much more adaptable
system that can generalize beyond intra-dataset samples, but to inter-dataset
samples, which it has never seen before during training. Quantitatively, our
method approaches the baseline performance of the current state of the art
anti-spoofing models with 15.8x less parameters used. Moreover, experiments
showed that the introduced methodology performs well only using basic binary
label without additional semantic information which indicates potential
benefits of this work in industrial and application based environment where
trade-off between additional labelling and resources are considered.
|
A transient two-dimensional acoustic boundary element solver is coupled to a
potential flow boundary element solver via Powell's acoustic analogy to
determine the acoustic emission of isolated hydrofoils performing
biologically-inspired motions. The flow-acoustic boundary element framework is
validated against experimental and asymptotic solutions for the noise produced
by canonical vortex-body interactions. The numerical framework then
characterizes the noise production of an oscillating foil, which is a simple
representation of a fish caudal fin. A rigid NACA 0012 hydrofoil is subjected
to combined heaving and pitching motions for Strouhal numbers ($0.03 < St < 1$)
based on peak-to-peak amplitudes and chord-based reduced frequencies ($0.125 <
f^* < 1$) that span the parameter space of many swimming fish species. A
dipolar acoustic directivity is found for all motions, frequencies, and
amplitudes considered, and the peak noise level increases with both the reduced
frequency and the Strouhal number. A combined heaving and pitching motion
produces less noise than either a purely pitching or purely heaving foil at a
fixed reduced frequency and amplitude of motion. Correlations of the lift and
power coefficients with the peak root-mean-square acoustic pressure levels are
determined, which could be utilized to develop long-range, quiet swimmers.
|
In power system dynamic simulation, up to 90% of the computational time is
devoted to solve the network equations, i.e., a set of linear equations.
Traditional approaches are based on sparse LU factorization, which is
inherently sequential. In this paper, an inverse-based network solution is
proposed by a hierarchical method for computing and store the approximate
inverse of the conductance matrix in electromagnetic transient (EMT)
simulations. The proposed method can also efficiently update the inverse by
modifying only local sub-matrices to reflect changes in the network, e.g., loss
of a line. Experiments on a series of simplified 179-bus Western
Interconnection demonstrate the advantages of the proposed methods.
|
We present the results of long-term photometric monitoring of two active
galactic nuclei, 2MASX J08535955+7700543 (z $\sim$ 0.106) and VII Zw 244 (z
$\sim$ 0.131), being investigated by the reverberation mapping method in
medium-band filters. To estimate the size of the broad line region, we have
analyzed the light curves with the JAVELIN code. The emission line widths have
been measured using the spectroscopic data obtained at the 6-m BTA telescope of
SAO RAS. We give our estimates of the supermassive black hole masses $\lg
(M/M_{\odot})$, $7.398_{-0.171}^{+0.153}$, and $7.049_{-0.075}^{+0.068}$,
respectively
|
Perpendicularly magnetized films showing small saturation magnetization,
$M_\mathrm{s}$, are essential for spin-transfer-torque writing type
magnetoresistive random access memories, STT-MRAMs. An intermetallic compound,
{(Mn-Cr)AlGe} of the Cu$_2$Sb-type crystal structure was investigated, in this
study, as a material showing the low $M_\mathrm{s}$ ($\sim 300$ kA/m) and
high-perpendicular magnetic anisotropy, $K_\mathrm{u}$. The layer thickness
dependence of $K_\mathrm{u}$ and effects of Mg-insertion layers at top and
bottom (Mn-Cr)AlGe$|$MgO interfaces were studied in film samples fabricated
onto thermally oxidized silicon substrates to realize high-$K_\mathrm{u}$ in
the thickness range of a few nanometer. Optimum Mg-insertion thicknesses were
1.4 and 3.0 nm for the bottom and the top interfaces, respectively, which were
relatively thick compared to results in similar insertion effect investigations
on magnetic tunnel junctions reported in previous studies. The cross-sectional
transmission electron microscope images revealed that the Mg-insertion layers
acted as barriers to interdiffusion of Al-atoms as well as oxidization from the
MgO layers. The values of $K_\mathrm{u}$ were about $7 \times 10^5$ and $2
\times 10^5$ J/m$^3$ at room temperature for 5 and 3 nm-thick (Mn-Cr)AlGe
films, respectively, with the optimum Mg-insertion thicknesses. The
$K_\mathrm{u}$ at a few nanometer thicknesses is comparable or higher than
those reported in perpendicularly magnetized CoFeB films which are
conventionally used in MRAMs, while the $M_\mathrm{s}$ value is one third or
less smaller than those of the CoFeB films. The developed (Mn-Cr)AlGe films are
promising from the viewpoint of not only the magnetic properties, but also the
compatibility to the silicon process in the film fabrication.
|
Blockchain (BC) technology can revolutionize future networks by providing a
distributed, secure, and unalterable way to boost collaboration among
operators, users, and other stakeholders. Its implementations have
traditionally been supported by wired communications, with performance
indicators like the high latency introduced by the BC being one of the key
technology drawbacks. However, when applied to wireless communications, the
performance of BC remains unknown, especially if running over contention-based
networks. In this paper, we evaluate the latency performance of BC technology
when the supporting communication platform is wireless, specifically we focus
on IEEE 802.11ax, for the use case of users' radio resource provisioning. For
that purpose, we propose a discrete-time Markov model to capture the expected
delay incurred by the BC. Unlike other models in the literature, we consider
the effect that timers and forks have on the end-to-end latency.
|
Recently, Martinez-Penas and Kschischang (IEEE Trans. Inf. Theory, 2019)
showed that lifted linearized Reed-Solomon codes are suitable codes for error
control in multishot network coding. We show how to construct and decode lifted
interleaved linearized Reed-Solomon codes. Compared to the construction by
Martinez-Penas-Kschischang, interleaving allows to increase the decoding region
significantly (especially w.r.t. the number of insertions) and decreases the
overhead due to the lifting (i.e., increases the code rate), at the cost of an
increased packet size. The proposed decoder is a list decoder that can also be
interpreted as a probabilistic unique decoder. Although our best upper bound on
the list size is exponential, we present a heuristic argument and simulation
results that indicate that the list size is in fact one for most channel
realizations up to the maximal decoding radius.
|
This paper presents a detailed investigation of FeCr-based quaternary Heusler
alloys. By using ultrasoft pseudopotential, electronic and magnetic properties
of the compounds are studied within the framework of Density Functional Theory
(DFT) by using the Quantum Espresso package. The thermodynamic, mechanical, and
dynamical stability of the compounds is established through the comprehensive
study of different mechanical parameters and phonon dispersion curves. The
meticulous study of elastic parameters such as bulk, Young's, shear moduli,
etc. is done to understand different mechanical properties. The FeCr-based
compounds containing also Yttrium are studied to redress the contradictory
electronic and magnetic properties observed in the literature. The interesting
properties like half-metallicity and spin-gapless semiconducting (SGS) behavior
are realized in the compounds under study.
|
We study several variants of the problem of moving a convex polytope $K$,
with $n$ edges, in three dimensions through a flat rectangular (and sometimes
more general) window. Specifically:
$\bullet$ We study variants where the motion is restricted to translations
only, discuss situations where such a motion can be reduced to sliding
(translation in a fixed direction), and present efficient algorithms for those
variants, which run in time close to $O(n^{8/3})$.
$\bullet$ We consider the case of a `gate' (an unbounded window with two
parallel infinite edges), and show that $K$ can pass through such a window, by
any collision-free rigid motion, if and only if it can slide through it.
$\bullet$ We consider arbitrary compact convex windows, and show that if $K$
can pass through such a window $W$ (by any motion) then $K$ can slide through a
gate of width equal to the diameter of $W$.
$\bullet$ We study the case of a circular window $W$, and show that, for the
regular tetrahedron $K$ of edge length $1$, there are two thresholds $1 >
\delta_1\approx 0.901388 > \delta_2\approx 0.895611$, such that (a) $K$ can
slide through $W$ if the diameter $d$ of $W$ is $\ge 1$, (b) $K$ cannot slide
through $W$ but can pass through it by a purely translational motion when
$\delta_1\le d < 1$, (c) $K$ cannot pass through $W$ by a purely translational
motion but can do it when rotations are allowed when $\delta_2 \le d <
\delta_1$, and (d) $K$ cannot pass through $W$ at all when $d < \delta_2$.
$\bullet$ Finally, we explore the general setup, where we want to plan a
general motion (with all six degrees of freedom) for $K$ through a rectangular
window $W$, and present an efficient algorithm for this problem, with running
time close to $O(n^4)$.
|
The paper is devoted to the study of Gromov-Hausdorff convergence and
stability of irreversible metric-measure spaces, both in the compact and
noncompact cases. While the compact setting is mostly similar to the reversible
case developed by J. Lott, K.-T. Sturm and C. Villani, the noncompact case
provides various surprising phenomena. Since the reversibility of noncompact
irreversible spaces might be infinite, it is motivated to introduce a suitable
nondecreasing function that bounds the reversibility of larger and larger
balls. By this approach, we are able to prove satisfactory
convergence/stability results in a suitable -- reversibility depending --
Gromov-Hausdorff topology. A wide class of irreversible spaces is provided by
Finsler manifolds, which serve to construct various model examples by pointing
out genuine differences between the reversible and irreversible settings. We
conclude the paper by proving various geometric and functional inequalities (as
Brunn-Minkowski, Bishop-Gromov, log-Sobolev and Lichnerowicz inequalities) on
irreversible structures.
|
We introduce an evolutionary game on hypergraphs in which decisions between a
risky alternative and a safe one are taken in social groups of different sizes.
The model naturally reproduces choice shifts, namely the differences between
the preference of individual decision makers and the consensual choice of a
group, that have been empirically observed in choice dilemmas. In particular, a
deviation from the Nash equilibrium towards the risky strategy occurs when the
dynamics takes place on heterogeneous hypergraphs. These results can explain
the emergence of irrational herding and radical behaviours in social groups.
|
Photos of faces captured in unconstrained environments, such as large crowds,
still constitute challenges for current face recognition approaches as often
faces are occluded by objects or people in the foreground. However, few studies
have addressed the task of recognizing partial faces. In this paper, we propose
a novel approach to partial face recognition capable of recognizing faces with
different occluded areas. We achieve this by combining attentional pooling of a
ResNet's intermediate feature maps with a separate aggregation module. We
further adapt common losses to partial faces in order to ensure that the
attention maps are diverse and handle occluded parts. Our thorough analysis
demonstrates that we outperform all baselines under multiple benchmark
protocols, including naturally and synthetically occluded partial faces. This
suggests that our method successfully focuses on the relevant parts of the
occluded face.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.