id
stringlengths 9
16
| abstract
stringlengths 67
2.61k
| cats
sequence | primary
stringlengths 5
18
| secondary
stringlengths 0
18
| strlabel
stringlengths 5
315
| stratlabel
class label 7.27k
classes |
|---|---|---|---|---|---|---|
2204.06232 | The dynamics of the partially ionized solar atmosphere is controlled by the
frequent collision and charge exchange between the predominant neutral Hydrogen
atoms and charged ions. At signal frequencies below or of the order of either
of the collision or charge exchange frequencies the magnetic stress is {\it
felt} by both the charged and neutral particles simultaneously. The resulting
neutral-mass loading of the ions leads to the rescaling of the effective
ion-cyclotron frequency-it becomes the Hall frequency, and the resultant
effective Larmor radius becomes of the order of few kms. Thus the finite Larmor
radius (FLR) effect which manifests as the ion and neutral pressure stress
tensors operates over macroscopic scales. Whereas parallel and perpendicular
(with respect to the magnetic field) viscous momentum transport competes with
the Ohm and Hall diffusion of the magnetic field in the
photosphere-chromosphre, the gyroviscous effect becomes important only in the
transition region between the chromosphere and corona, where it competes with
the ambipolar diffusion. The wave propagation in the gyroviscous effect
dominated medium depends on the plasma $\beta$ (a ratio of the thermal and
magnetic energies). The abundance of free energy makes gyro waves unstable with
the onset condition exactly opposite of the Hall instability. However, the
maximum growth rate is identical to the Hall instability. For a flow gradient
$\sim 0.1 \,\mbox{s}^{-1}$ the instability growth time is one minute. Thus, the
transition region may become subject to this fast growing, gyroviscous
instability.
| [
"astro-ph.SR"
] | astro-ph.SR | Solar and Stellar Astrophysics | 6,668Solar and Stellar Astrophysics
|
|
1704.05822 | Maximum likelihood estimation (MLE) is one of the most important methods in
machine learning, and the expectation-maximization (EM) algorithm is often used
to obtain maximum likelihood estimates. However, EM heavily depends on initial
configurations and fails to find the global optimum. On the other hand, in the
field of physics, quantum annealing (QA) was proposed as a novel optimization
approach. Motivated by QA, we propose a quantum annealing extension of EM,
which we call the deterministic quantum annealing expectation-maximization
(DQAEM) algorithm. We also discuss its advantage in terms of the path integral
formulation. Furthermore, by employing numerical simulations, we illustrate how
it works in MLE and show that DQAEM outperforms EM.
| [
"stat.ML",
"cond-mat.stat-mech",
"physics.comp-ph",
"quant-ph"
] | stat.ML | cond-mat.stat-mech | Machine Learning;Statistical Mechanics;Computational Physics;Quantum Physics | 7,267longtail
|
astro-ph/0305394 | We have carried out a high resolution spectroscopic survey of the 220-250 Myr
old cluster NGC 6475: our main purpose is to investigate Li evolution during
the early stages of the Main Sequence. We have determined Li abundances for 33
late F to K-type X-ray selected cluster candidates, extending the samples
already available in the literature; for part of the stars we obtained radial
and rotational velocities, allowing us to confirm the membership and to check
for binarity. We also estimated the cluster metallicity which turned out to be
over-solar ([Fe/H]=+0.14 +/-0.06). Our Li analysis evidenced that (i) late
F-type stars (Teff > 6000 K) undergo a very small amount of Li depletion during
the early phases on the ZAMS; (ii) G-type stars (6000 > Teff > 5500 K) instead
do deplete lithium soon after arrival on the ZAMS. Whereas this result is not
new, we show that the time scale for Li depletion in these stars is almost
constant between 100 and 600 Myr; (iii) we confirm that the spread observed in
early K-type stars in younger clusters has converged by 220 Myr. No constraints
can be put on later-type stars. (iv) Finally, we investigate the effect of
metallicity on Li depletion by comparing NGC 6475 with the similar age cluster
M 34, but we show that the issue remains open, given the uncertain metallicity
of the latter cluster. By using the combined NGC 6475 + M 34 sample together
with the Hyades and the Pleiades, we compare quantitatively Li evolution from
the ZAMS to 600 Myr with theoretical predictions of standard models.
| [
"astro-ph"
] | astro-ph | Astrophysics | 463Astrophysics
|
|
2204.05708 | Recent experiments have shown that CeH9 and (Ce,La)H9 can be synthesized
under high pressure between 90-170GPa and become a superconductor with a high
value of superconducting critical temperature (Tc) between 100-200K. In this
work, we performed a theoretical study of a (Ce,La)H9 compound where the Ce:La
ratio is equal to 1. We used the density functional theory and the {\it ab
initio} molecular dynamics (AIMD) method. From the phonon dispersion, there
exist some unstable modes around the K-point phonons. Then, we performed AIMD
simulation at around 203K and found that the compound becomes stable. The
superconducting spectral function can be calculated. We found that $\lambda$ is
as high as 3.0 at 200GPa. By using Allen-Dynes-modified McMillan equation in
the strong coupling regime, we found that Tc = 87K at 200GPa.
| [
"cond-mat.supr-con",
"physics.comp-ph"
] | cond-mat.supr-con | physics.comp-ph | Superconductivity;Computational Physics | 7,069Superconductivity;Computational Physics
|
1209.0789 | The corona of the Sun is dominated by emission from loop-like structures.
When observed in X-ray or extreme ultraviolet emission, these million K hot
coronal loops show a more or less constant cross section. In this study we show
how the interplay of heating, radiative cooling, and heat conduction in an
expanding magnetic structure can explain the observed constant cross section.
We employ a three-dimensional magnetohydrodynamics (3D MHD) model of the
corona. The heating of the coronal plasma is the result of braiding of the
magnetic field lines through footpoint motions and subsequent dissipation of
the induced currents. From the model we synthesize the coronal emission, which
is directly comparable to observations from, e.g., the Atmospheric Imaging
Assembly on the Solar Dynamics Observatory (AIA/SDO). We find that the
synthesized observation of a coronal loop seen in the 3D data cube does match
actually observed loops in count rate and that the cross section is roughly
constant, as observed. The magnetic field in the loop is expanding and the
plasma density is concentrated in this expanding loop; however, the temperature
is not constant perpendicular to the plasma loop. The higher temperature in the
upper outer parts of the loop is so high that this part of the loop is outside
the contribution function of the respective emission line(s). In effect, the
upper part of the plasma loop is not bright and thus the loop actually seen in
coronal emission appears to have a constant width. From this we can conclude
that the underlying field-line-braiding heating mechanism provides the proper
spatial and temporal distribution of the energy input into the corona --- at
least on the observable scales.
| [
"astro-ph.SR"
] | astro-ph.SR | Solar and Stellar Astrophysics | 6,668Solar and Stellar Astrophysics
|
|
2301.11531 | We present a quantum annealing-based solution method for topology
optimization (TO). In particular, we consider TO in a more general setting,
i.e., applied to structures of continuum domains where designs are represented
as distributed functions, referred to as continuum TO problems. According to
the problem's properties and structure, we formulate appropriate sub-problems
that can be solved on an annealing-based quantum computer. The methodology
established can effectively tackle continuum TO problems formulated as
mixed-integer nonlinear programs. To maintain the resulting sub-problems small
enough to be solved on quantum computers currently accessible with small
numbers of quits and limited connectivity, we further develop a splitting
approach that splits the problem into two parts: the first part can be
efficiently solved on classical computers, and the second part with a reduced
number of variables is solved on a quantum computer. By such, a practical
continuum TO problem of varying scales can be handled on the D-Wave quantum
annealer. More specifically, we concern the minimum compliance, a canonical TO
problem that seeks an optimal distribution of materials to minimize the
compliance with desired material usage. The superior performance of the
developed methodology is assessed and compared with the state-of-the-art
heuristic classical methods, in terms of both solution quality and
computational efficiency. The present work hence provides a promising new
avenue of applying quantum computing to practical designs of topology for
various applications.
| [
"math.NA",
"cs.CE",
"cs.NA",
"math.OC"
] | math.NA | cs.CE | Numerical Analysis;Computational Engineering, Finance, and Science;Numerical Analysis;Optimization and Control | 7,267longtail
|
2208.13362 | Krylov complexity is a novel observable for detecting quantum chaos, and an
indicator of a possible gravity dual. In this paper, we compute the Krylov
complexity and the associated Lanczos coefficients in the SU(2) Yang-Mills
theory, which can be reduced to a nonlinearly coupled harmonic oscillators
(CHO) model. We show that there exists a chaotic transition in the growth of
Krylov complexity. The Krylov complexity shows a quadratic growth in the early
time stage and then grows linearly. The corresponding Lanczos coefficient
satisfies the universal operator growth hypothesis, i.e., grows linearly first
and then enters the saturation plateau. By the linear growth of Lanczos
coefficients, we obtain an upper bound of the quantum Lyapunov exponent.
Finally, we investigate the effect of different energy sectors on the
K-complexity and Lanczos coefficients.
| [
"hep-th"
] | hep-th | High Energy Physics - Theory | 3,266High Energy Physics - Theory
|
|
math/0209050 | Various best-choice problems related to the planar homogeneous Poisson
process in finite or semi-infinite rectangle are studied. The analysis is
largely based on properties of the one-dimensional box-area process associated
with the sequence of records. We prove a series of distributional identities
involving exponential and uniform random variables, and resolve the
Petruccelli-Porosinski-Samuels paradox on coincidence of asymptotic values in
certain discrete-time optimal stopping problems.
| [
"math.PR"
] | math.PR | Probability | 5,709Probability
|
|
1703.02827 | In this paper, the containment problem for the defining ideal of a special
type of zero dimensional subschemes of $\mathbb{P}^2$, so called quasi star
configurations, is investigated. Some sharp bounds for the resurgence of these
types of ideals are given. As an application of this result, for every real
number $0 < \varepsilon < \frac{1}{2}$, we construct an infinite family of
homogeneous radical ideals of points in $\mathbb{K}[\mathbb{P}^2]$ such that
their resurgences lie in the interval $[2- \varepsilon ,2)$. Moreover, the
Castelnuovo-Mumford regularity of all ordinary powers of defining ideal of
quasi star configurations are determined. In particular, it is shown that, %the
defining ideal of a quasi star configuration, and all of them have linear
resolution.
| [
"math.AG"
] | math.AG | Algebraic Geometry | 47Algebraic Geometry
|
|
1905.05666 | A few years ago the use of standard functional manipulations was demonstrated
to imply an unexpected property satisfied by the fermionic Green's functions of
QCD, and called effective locality. This feature of QCD is non-perturbative as
it results from a full integration of the gluonic degrees of freedom. In this
paper, at eikonal and quenching approximation at least, the relation of
effective locality to dynamical chiral symmetry breaking is examined.
| [
"hep-th"
] | hep-th | High Energy Physics - Theory | 3,266High Energy Physics - Theory
|
|
hep-th/9701076 | An earlier proposed theory with linear-gonihedhic action for quantum gravity
is reviewed. One can consider this theory as a "square root" of classical
gravity with a new fundamental constant of dimension one. We demonstrate also,
that the partition function for the discretized version of the Einstein-Hilbert
action found by Regge in 1961 can be represented as a superposition of random
surfaces with Euler character as an action and in the case of linear gravity as
a superposition of three-dimensional manifolds with an action which is
proportional to the total solid angle deficit of these manifolds. This
representation allows to construct the transfer matrix which describes the
propagation of space manifold. We discuss the so called gonihedric principle
which allows to defind a discrete version of high derivative terms in quantum
gravity and to introduce intrinsic rigidity of spacetime. This note is based on
a talk delivered at the II meeting on constrained dynamics and quantum gravity
at Santa Margherita Ligure.
| [
"hep-th"
] | hep-th | High Energy Physics - Theory | 3,266High Energy Physics - Theory
|
|
1107.3657 | By considering a continuous pruning procedure on Aldous's Brownian tree, we
construct a random variable $\Theta$ which is distributed, conditionally given
the tree, according to the probability law introduced by Janson as the limit
distribution of the number of cuts needed to isolate the root in a critical
Galton-Watson tree. We also prove that this random variable can be obtained as
the a.s. limit of the number of cuts needed to cut down the subtree of the
continuum tree spanned by $n$ leaves.
| [
"math.PR"
] | math.PR | Probability | 5,709Probability
|
|
hep-ph/9609311 | Measurements of distributions associated with the pair production of top
quarks at the LHC can be used to constrain (or observe) the anomalous
chromomagnetic dipole moment($\kappa$) of the top. For example, using either
the $t\bar t$ invariant mass or the $p_t$ distribution of top we find that
sensitivities to $|\kappa|$ of order 0.05 are obtainable with 100 $fb^{-1}$ of
integrated luminosity. This is similar in magnitude to what can be obtained at
a 500 GeV NLC with an integrated luminosity of 50 $fb^{-1}$ through an
examination of the $e^+e^- \to t\bar tg$ process. [To appear in the Proceedings
of the 1996 DPF/DPB Summer Study on NewDirections for High Energy
Physics-Snowmass96, Snowmass, CO, 25 June-12 July, 1996.]
| [
"hep-ph"
] | hep-ph | High Energy Physics - Phenomenology | 3,129High Energy Physics - Phenomenology
|
|
2310.17064 | As artificial intelligence (AI) gains greater adoption in a wide variety of
applications, it has immense potential to contribute to mathematical discovery,
by guiding conjecture generation, constructing counterexamples, assisting in
formalizing mathematics, and discovering connections between different
mathematical areas, to name a few.
While prior work has leveraged computers for exhaustive mathematical proof
search, recent efforts based on large language models (LLMs) aspire to position
computing platforms as co-contributors in the mathematical research process.
Despite their current limitations in logic and mathematical tasks, there is
growing interest in melding theorem proving systems with foundation models.
This work investigates the applicability of LLMs in formalizing advanced
mathematical concepts and proposes a framework that can critically review and
check mathematical reasoning in research papers. Given the noted reasoning
shortcomings of LLMs, our approach synergizes the capabilities of proof
assistants, specifically PVS, with LLMs, enabling a bridge between textual
descriptions in academic papers and formal specifications in PVS. By harnessing
the PVS environment, coupled with data ingestion and conversion mechanisms, we
envision an automated process, called \emph{math-PVS}, to extract and formalize
mathematical theorems from research papers, offering an innovative tool for
academic review and discovery.
| [
"cs.AI",
"cs.CL",
"cs.LG",
"cs.LO"
] | cs.AI | cs.CL | Artificial Intelligence;Computation and Language;Machine Learning;Logic in Computer Science | 7,267longtail
|
1508.00632 | We show how to price and replicate a variety of barrier-style claims written
on the $\log$ price $X$ and quadratic variation $\langle X \rangle$ of a risky
asset. Our framework assumes no arbitrage, frictionless markets and zero
interest rates. We model the risky asset as a strictly positive continuous
semimartingale with an independent volatility process. The volatility process
may exhibit jumps and may be non-Markovian. As hedging instruments, we use only
the underlying risky asset, zero-coupon bonds, and European calls and puts with
the same maturity as the barrier-style claim. We consider knock-in, knock-out
and rebate claims in single and double barrier varieties.
| [
"q-fin.MF"
] | q-fin.MF | Mathematical Finance | 4,385Mathematical Finance
|
|
cond-mat/0111217 | Eu2-xCexRuSr2Cu2O10-d (Ru-2122) is the first Cu-O based system in which
superconductivity (SC) in the CuO2 planes and weak-ferromagnetism (W-FM) in the
Ru sub-lattice coexists. The hole doping in the CuO2 planes, is controlled by
appropriate variation of the Ce concentration. SC occurs for Ce contents of
0.4-0.8, with the highest TC=35 K for Ce=0.6. The as-prepared non-SC
EuCeRuSr2Cu2O10 (x=1) sample exhibits magnetic irreversibility below Tirr=125 K
and orders anti-ferromagnetically (AFM) at TM =165 K. The saturation moment at
5 K is Msat=0.89 mB /Ru close to the expected 1 mB for the low-spin state of
Ru5+. Annealing under oxygen pressures, does not affect these parameters,
whereas depletion of oxygen shifts both Tirr and TM up to 169 and 215 K
respectively. Systematic magnetic studies on Eu2-xCexRuSr2Cu2O10-d show that
TM, Tirr and Msat decrease with x, and the Ce dependent magnetic-SC phase
diagram is presented. A simple model for the SC state is proposed. We interpret
the magnetic behavior in the framework of our ac and dc magnetic studies, and
argue that: (i) the system becomes AFM ordered at TM; (b) at Tirr < TM, W-FM is
induced by the canting of the Ru moments, and (c), at lower temperatures the
appropriate samples become SC at TC. The magnetic features are not affected by
the SC state, and the two states coexist.
| [
"cond-mat.supr-con"
] | cond-mat.supr-con | Superconductivity | 7,066Superconductivity
|
|
1903.03454 | The negative hydrogen ion is the first three body quantum problem whose
ground state energy is calculated using the `Chandrasekhar Wavefunction' that
accounts for the electron-electron correlation. Solving multi-body systems is a
daunting task in quantum mechanics as it includes choosing a trial wavefunction
and the calculation of integrals for the system that becomes almost impossible
for systems with three or more particles. This difficulty can be addressed by
quantum computers. They have emerged as a tool to address different electronic
structure problems with remarkable efficiency. They have been realized in
various fields and proved their efficiency over classical computers. Here, we
show the quantum simulation of H^{-} ion to calculate it's ground state energy
in IBM quantum computer. The energy is found to be -0.5339355468 Hartree with
an error of 0.8376% as compared to the theoretical value. We observe that the
quantum computer is efficient in preparing the correlated wavefunction of H^{-}
and calculating it's ground state energy. We use a recently developed algorithm
known as `Variational Quantum Eigensolver' and implement it in IBM's 5-qubit
quantum chip `ibmqx2'. The method consists of a quantum part i.e., state
preparation and measurement of expectation values using the quantum computer,
and the classical part i.e., the optimization routine run in a classical
computer for energy convergence. An optimization routine is performed on
classical computer by running quantum chemistry program and codes in QISKit to
converge the energy to the minimum. We also present a comparison of different
optimization routines and encoding methods used to converge the energy value to
the minimum. The technique can be used to solve various many body problems with
great efficiency.
| [
"quant-ph"
] | quant-ph | Quantum Physics | 5,985Quantum Physics
|
|
astro-ph/0101561 | A LiBeB evolution model including Galactic Cosmic Ray nucleosynthesis, the
$\nu$-process, novae, AGB and C-stars is presented. We have included Galactic
Cosmic Ray Nucleosynthesis (GCRN) in a complete Chemical Evolution Model that
takes into account 76 stable isotopes from hydrogen to zinc. Any successful
LiBeB evolution model should also be compatible with other observational
constraints like the age-metallicity relation, the G-dwarf distribution or the
evolution of other elements. At the same time, we have checked how different
would be a model that took into account the last observations by Wakker et al.
(1999) of metal-enriched clouds falling onto the disk, from a primordial infall
model.
| [
"astro-ph"
] | astro-ph | Astrophysics | 463Astrophysics
|
|
1707.02676 | In hierarchical searches for continuous gravitational waves, clustering of
candidates is an important postprocessing step because it reduces the number of
noise candidates that are followed-up at successive stages [1][7][12]. Previous
clustering procedures bundled together nearby candidates ascribing them to the
same root cause (be it a signal or a disturbance), based on a predefined
cluster volume. In this paper, we present a procedure that adapts the cluster
volume to the data itself and checks for consistency of such volume with what
is expected from a signal. This significantly improves the noise rejection
capabilities at fixed detection threshold, and at fixed computing resources for
the follow-up stages, this results in an overall more sensitive search. This
new procedure was employed in the first Einstein@Home search on data from the
first science run of the advanced LIGO detectors (O1) [11].
| [
"gr-qc",
"astro-ph.IM",
"math.GN"
] | gr-qc | astro-ph.IM | General Relativity and Quantum Cosmology;Instrumentation and Methods for Astrophysics;General Topology | 7,267longtail
|
1309.4170 | Measurements in the radio regime embrace a number of effective approaches for
WISP searches, often covering unique or highly complementary ranges of the
parameter space compared to those explored in other research domains. These
measurements can be used to search for electromagnetic tracers of the hidden
photon and axion oscillations, extending down to ~10^-19 eV the range of the
hidden photon mass probed, and closing the last gaps in the strongly favoured
1-5 micro-eV range for axion dark matter. This provides a strong impetus for
several new initiatives in the field, including the WISP Dark Matter eXperiment
(WISPDMX) and novel conceptual approaches for broad-band WISP searches in the
0.1-1000 micro-eV range.
| [
"physics.ins-det",
"astro-ph.CO",
"hep-ex",
"hep-ph"
] | physics.ins-det | astro-ph.CO | Instrumentation and Detectors;Cosmology and Nongalactic Astrophysics;High Energy Physics - Experiment;High Energy Physics - Phenomenology | 3,639Instrumentation and Detectors;Cosmology and Nongalactic Astrophysics;High Energy Physics - Experiment;High Energy Physics - Phenomenology
|
2104.10900 | This paper presents a novel preconditioning strategy for the classic 8-point
algorithm (8-PA) for estimating an essential matrix from 360-FoV images (i.e.,
equirectangular images) in spherical projection. To alleviate the effect of
uneven key-feature distributions and outlier correspondences, which can
potentially decrease the accuracy of an essential matrix, our method optimizes
a non-rigid transformation to deform a spherical camera into a new spatial
domain, defining a new constraint and a more robust and accurate solution for
an essential matrix. Through several experiments using random synthetic points,
360-FoV, and fish-eye images, we demonstrate that our normalization can
increase the camera pose accuracy by about 20% without significantly overhead
the computation time. In addition, we present further benefits of our method
through both a constant weighted least-square optimization that improves
further the well known Gold Standard Method (GSM) (i.e., the non-linear
optimization by using epipolar errors); and a relaxation of the number of
RANSAC iterations, both showing that our normalization outcomes a more
reliable, robust, and accurate solution.
| [
"cs.CV",
"cs.RO"
] | cs.CV | cs.RO | Computer Vision and Pattern Recognition;Robotics | 1,636Computer Vision and Pattern Recognition;Robotics
|
2010.10352 | Moving loads such as cars and trains are very useful sources of seismic
waves, which can be analyzed to retrieve information on the seismic velocity of
subsurface materials using the techniques of ambient noise seismology. This
information is valuable for a variety of applications such as geotechnical
characterization of the near-surface, seismic hazard evaluation, and
groundwater monitoring. However, for such processes to converge quickly, data
segments with appropriate noise energy should be selected. Distributed Acoustic
Sensing (DAS) is a novel sensing technique that enables acquisition of these
data at very high spatial and temporal resolution for tens of kilometers. One
major challenge when utilizing the DAS technology is the large volume of data
that is produced, thereby presenting a significant Big Data challenge to find
regions of useful energy. In this work, we present a highly scalable and
efficient approach to process real, complex DAS data by integrating physics
knowledge acquired during a data exploration phase followed by deep supervised
learning to identify "useful" coherent surface waves generated by anthropogenic
activity, a class of seismic waves that is abundant on these recordings and is
useful for geophysical imaging. Data exploration and training were done on
130~Gigabytes (GB) of DAS measurements. Using parallel computing, we were able
to do inference on an additional 170~GB of data (or the equivalent of 10 days'
worth of recordings) in less than 30 minutes. Our method provides interpretable
patterns describing the interaction of ground-based human activities with the
buried sensors.
| [
"eess.SP",
"cs.LG",
"physics.geo-ph"
] | eess.SP | cs.LG | Signal Processing;Machine Learning;Geophysics | 7,267longtail
|
2106.15588 | A dessin d'enfant, or dessin, is a bicolored graph embedded into a Riemann
surface, and the monodromy group is an algebraic invariant of the dessin
generated by rotations of edges about black and white vertices. A rational
billiards surface is a two dimensional surface that allows one to view the path
of a billiards ball as a continuous path. In this paper, we classify the
monodromy groups of dessins associated to rational triangular billiards
surfaces.
| [
"math.NT"
] | math.NT | Number Theory | 4,945Number Theory
|
|
1403.7158 | We prove sharp inequalities for the average number of affine diameters
through the points of a convex body $K$ in ${\mathbb R}^n$. These inequalities
hold if $K$ is either a polytope or of dimension two. An example shows that the
proof given in the latter case does not extend to higher dimensions.
| [
"math.MG"
] | math.MG | Metric Geometry | 4,601Metric Geometry
|
|
hep-ex/0408129 | We report a study of the suppressed decay $B^{-} \to
[K^{+}\pi^{-}]_{D}K^{-}$(and its charge-conjugate mode) at Belle, where
$[K^{+}\pi^{-}]_{D}$ indicates that the $K^{+}\pi^{-}$ pair originates from a
neutral $D$ meson. A data sample containing 274 million $B\bar{B}$ pairs
recorded at the $\Upsilon(4S)$ resonance with the Belle detector at the KEKB
asymmetric $e^{+}e^{-}$ storage ring is used. This decay mode can be used to
extract the CKM angle $\phi_{3}$ using the so-called Atwood-Dunietz-Soni
method. The signal for $B^{-} \to [K^{+}\pi^{-}]_{D}K^{-}$ has $2.7\sigma$
statistical significance, and we set a limit on the ratio of B decay amplitudes
$r_B < 0.28$ at the 90% confidence level. We observe a signal with $5.8\sigma$
statistical significance in the related mode, $B^{-} \to
[K^{+}\pi^{-}]_{D}\pi^{-}$.
| [
"hep-ex"
] | hep-ex | High Energy Physics - Experiment | 3,059High Energy Physics - Experiment
|
|
1802.09900 | To launch black-box attacks against a Deep Neural Network (DNN) based Face
Recognition (FR) system, one needs to build \textit{substitute} models to
simulate the target model, so the adversarial examples discovered from
substitute models could also mislead the target model. Such
\textit{transferability} is achieved in recent studies through querying the
target model to obtain data for training the substitute models. A real-world
target, likes the FR system of law enforcement, however, is less accessible to
the adversary. To attack such a system, a substitute model with similar quality
as the target model is needed to identify their common defects. This is hard
since the adversary often does not have the enough resources to train such a
powerful model (hundreds of millions of images and rooms of GPUs are needed to
train a commercial FR system).
We found in our research, however, that a resource-constrained adversary
could still effectively approximate the target model's capability to recognize
\textit{specific} individuals, by training \textit{biased} substitute models on
additional images of those victims whose identities the attacker want to cover
or impersonate. This is made possible by a new property we discovered, called
\textit{Nearly Local Linearity} (NLL), which models the observation that an
ideal DNN model produces the image representations (embeddings) whose distances
among themselves truthfully describe the human perception of the differences
among the input images. By simulating this property around the victim's images,
we significantly improve the transferability of black-box impersonation attacks
by nearly 50\%. Particularly, we successfully attacked a commercial system
trained over 20 million images, using 4 million images and 1/5 of the training
time but achieving 62\% transferability in an impersonation attack and 89\% in
a dodging attack.
| [
"cs.LG",
"cs.CV"
] | cs.LG | cs.CV | Machine Learning;Computer Vision and Pattern Recognition | 4,045Machine Learning;Computer Vision and Pattern Recognition
|
cond-mat/0204590 | One dimensional spin-1/2 $XXZ$ model in a transverse magnetic field is
studied. It is shown that the field induces the gap in the spectrum of the
model with easy-plain anisotropy. Using conformal invariance the field
dependence of the gap at small fields is found. The ground state phase diagram
is obtained. It contains four phases with different types of the long range
order (LRO) and a disordered one. These phases are separated by critical lines,
where the gap and the long range order vanish. Using scaling estimations and a
mean-field approach as well as numerical calculations in the vicinity of all
critical lines we found the critical exponents of the gap and the LRO. It is
shown that transition line between the ordered and disordered phases belongs to
the universality class of the transverse Ising model.
| [
"cond-mat.str-el"
] | cond-mat.str-el | Strongly Correlated Electrons | 6,979Strongly Correlated Electrons
|
|
1403.1142 | Security APIs, key servers and protocols that need to keep the status of
transactions, require to maintain a global, non-monotonic state, e.g., in the
form of a database or register. However, most existing automated verification
tools do not support the analysis of such stateful security protocols -
sometimes because of fundamental reasons, such as the encoding of the protocol
as Horn clauses, which are inherently monotonic. A notable exception is the
recent tamarin prover which allows specifying protocols as multiset rewrite
(msr) rules, a formalism expressive enough to encode state. As multiset
rewriting is a "low-level" specification language with no direct support for
concurrent message passing, encoding protocols correctly is a difficult and
error-prone process. We propose a process calculus which is a variant of the
applied pi calculus with constructs for manipulation of a global state by
processes running in parallel. We show that this language can be translated to
msr rules whilst preserving all security properties expressible in a dedicated
first-order logic for security properties. The translation has been implemented
in a prototype tool which uses the tamarin prover as a backend. We apply the
tool to several case studies among which a simplified fragment of PKCS\#11, the
Yubikey security token, and an optimistic contract signing protocol.
| [
"cs.CR"
] | cs.CR | Cryptography and Security | 1,782Cryptography and Security
|
|
2112.10070 | So far, named entity recognition (NER) has been involved with three major
types, including flat, overlapped (aka. nested), and discontinuous NER, which
have mostly been studied individually. Recently, a growing interest has been
built for unified NER, tackling the above three jobs concurrently with one
single model. Current best-performing methods mainly include span-based and
sequence-to-sequence models, where unfortunately the former merely focus on
boundary identification and the latter may suffer from exposure bias. In this
work, we present a novel alternative by modeling the unified NER as word-word
relation classification, namely W^2NER. The architecture resolves the kernel
bottleneck of unified NER by effectively modeling the neighboring relations
between entity words with Next-Neighboring-Word (NNW) and Tail-Head-Word-*
(THW-*) relations. Based on the W^2NER scheme we develop a neural framework, in
which the unified NER is modeled as a 2D grid of word pairs. We then propose
multi-granularity 2D convolutions for better refining the grid representations.
Finally, a co-predictor is used to sufficiently reason the word-word relations.
We perform extensive experiments on 14 widely-used benchmark datasets for flat,
overlapped, and discontinuous NER (8 English and 6 Chinese datasets), where our
model beats all the current top-performing baselines, pushing the
state-of-the-art performances of unified NER.
| [
"cs.CL"
] | cs.CL | Computation and Language | 1,168Computation and Language
|
|
1811.02668 | Recent studies have shown promising results in using Deep Learning to detect
malignancy in whole slide imaging. However, they were limited to just
predicting positive or negative finding for a specific neoplasm. We attempted
to use Deep Learning with a convolutional neural network algorithm to build a
lymphoma diagnostic model for four diagnostic categories: benign lymph node,
diffuse large B cell lymphoma, Burkitt lymphoma, and small lymphocytic
lymphoma. Our software was written in Python language. We obtained digital
whole slide images of Hematoxylin and Eosin stained slides of 128 cases
including 32 cases for each diagnostic category. Four sets of 5 representative
images, 40x40 pixels in dimension, were taken for each case. A total of 2,560
images were obtained from which 1,856 were used for training, 464 for
validation and 240 for testing. For each test set of 5 images, the predicted
diagnosis was combined from prediction of 5 images. The test results showed
excellent diagnostic accuracy at 95% for image-by-image prediction and at 10%
for set-by-set prediction. This preliminary study provided a proof of concept
for incorporating automated lymphoma diagnostic screen into future pathology
workflow to augment the pathologists' productivity.
| [
"cs.CV",
"cs.LG",
"stat.ML"
] | cs.CV | cs.LG | Computer Vision and Pattern Recognition;Machine Learning;Machine Learning | 1,601Computer Vision and Pattern Recognition;Machine Learning;Machine Learning
|
1507.02297 | We introduce new methods of equivalence checking and simulation based on
Computing Range Reduction (CRR). Given a combinational circuit $N$, the CRR
problem is to compute the set of outputs that disappear from the range of $N$
if a set of inputs of $N$ is excluded from consideration. Importantly, in many
cases, range reduction can be efficiently found even if computing the entire
range of $N$ is infeasible.
Solving equivalence checking by CRR facilitates generation of proofs of
equivalence that mimic a "cut propagation" approach. A limited version of such
an approach has been successfully used by commercial tools. Functional
verification of a circuit $N$ by simulation can be viewed as a way to reduce
the complexity of computing the range of $N$. Instead of finding the entire
range of $N$ and checking if it contains a bad output, such a range is computed
only for one input. Simulation by CRR offers an alternative way of coping with
the complexity of range computation. The idea is to exclude a subset of inputs
of $N$ and compute the range reduction caused by such an exclusion. If the set
of disappeared outputs contains a bad one, then $N$ is buggy.
| [
"cs.LO"
] | cs.LO | Logic in Computer Science | 3,801Logic in Computer Science
|
|
1110.3128 | In this paper, we consider constructibility of simplicial 3-balls. In many
cases, examining 1-dimensional subcomplexes of a simplicial 3-ball is efficient
to solve the decision problem whether the simplicial 3-ball is constructible or
not. From the point of view, we consider the case where a simplicial 3-ball has
spanning edges and present a sufficient condition for nonconstructibility.
| [
"math.CO"
] | math.CO | Combinatorics | 1,014Combinatorics
|
|
1501.04004 | We consider cosmological modelling in $f(R)$ theories of gravity, using both
top-down and bottom-up constructions. The top-down models are based on
Robertson-Walker geometries, and the bottom-up constructions are built by
patching together sub-horizon-sized regions of perturbed Minkowski space. Our
results suggest that these theories do not provide a theoretically attractive
alternative to the standard general relativistic cosmology. We find that the
only $f(R)$ theories that can admit an observationally viable weak-field limit
have large-scale expansions that are observationally indistinguishable from the
Friedmann solutions of General Relativity with $\Lambda$. Such theories do not
alleviate any of the difficulties associated with $\Lambda$, and cannot produce
any new behaviour in the cosmological expansion without simultaneously
destroying the Newtonian approximation to gravity on small scales.
| [
"gr-qc",
"astro-ph.CO"
] | gr-qc | astro-ph.CO | General Relativity and Quantum Cosmology;Cosmology and Nongalactic Astrophysics | 2,701General Relativity and Quantum Cosmology;Cosmology and Nongalactic Astrophysics
|
1702.05782 | The Principle of Maximum Entropy, a powerful and general method for inferring
the distribution function given a set of constraints, is applied to deduce the
overall distribution of 3D plasmoids (flux ropes/tubes) for systems where
resistive MHD is applicable and large numbers of plasmoids are produced. The
analysis is undertaken for the 3D case, with mass, total flux and velocity
serving as the variables of interest, on account of their physical and
observational relevance. The distribution functions for the mass, width, total
flux and helicity exhibit a power-law behavior with exponents of $-4/3$, $-2$,
$-3$ and $-2$ respectively for small values, whilst all of them display an
exponential falloff for large values. In contrast, the velocity distribution,
as a function of $v = |{\bf v}|$, is shown to be flat for $v \rightarrow 0$,
and becomes a power law with an exponent of $-7/3$ for $v \rightarrow \infty$.
Most of these results are nearly independent of the free parameters involved in
this specific problem. A preliminary comparison of our results with the
observational evidence is presented, and some of the ensuing space and
astrophysical implications are briefly discussed.
| [
"astro-ph.HE",
"astro-ph.SR",
"cond-mat.stat-mech",
"physics.plasm-ph",
"physics.space-ph"
] | astro-ph.HE | astro-ph.SR | High Energy Astrophysical Phenomena;Solar and Stellar Astrophysics;Statistical Mechanics;Plasma Physics;Space Physics | 7,267longtail
|
math/0011177 | Using the frame formalism we determine some possible metrics and
metric-compatible connections on the noncommutative differential geometry of
the real quantum plane. By definition a metric maps the tensor product of two
1-forms into a `function' on the quantum plane. It is symmetric in a modified
sense, namely in the definition of symmetry one has to replace the permutator
map with a deformed map \sigma fulfilling some suitable conditions.
Correspondingly, also the definition of the hermitean conjugate of the tensor
product of two 1-forms is modified (but reduces to the standard one if \sigma
coincides with the permutator). The metric is real with respect to such
modified *-structure.
| [
"math.QA"
] | math.QA | Quantum Algebra | 5,873Quantum Algebra
|
|
1710.10900 | Let $\vec{K}_{\mathbb{N}}$ be the complete symmetric digraph on the positive
integers. Answering a question of DeBiasio and McKenney, we construct a
2-colouring of the edges of $\vec{K}_{\mathbb{N}}$ in which every monochromatic
path has density 0. On the other hand, we show that, in every colouring that
does not have a directed path with $r$ edges in the first colour, there is
directed path in the second colour with density at least $\frac1r$.
| [
"math.CO"
] | math.CO | Combinatorics | 1,014Combinatorics
|
|
2209.04598 | This paper proposes a data-driven affinely adjustable robust Volt/VAr control
(AARVVC) scheme, which modulates the smart inverter reactive power in an affine
function of its active power, based on the voltage sensitivities with respect
to real/reactive power injections. To achieve a fast and accurate estimation of
voltage sensitivities, we propose a data-driven method based on deep neural
network (DNN), together with a rule-based bus-selection process using the
bidirectional search method. Our method only uses the operating statuses of
selected buses as inputs to DNN, thus significantly improving the training
efficiency and reducing information redundancy. Finally, a distributed
consensus-based solution, based on the alternating direction method of
multipliers (ADMM), for the AARVVC is applied to decide the inverter reactive
power adjustment rule with respect to its active power. Only limited
information exchange is required between each local agent and the central agent
to obtain the slope of the reactive power adjustment rule, and there is no need
for the central agent to solve any (sub)optimization problems. Numerical
results on the modified IEEE-123 bus system validate the effectiveness and
superiority of the proposed data-driven AARVVC method.
| [
"eess.SY",
"cs.SY"
] | eess.SY | cs.SY | Systems and Control;Systems and Control | 7,220Systems and Control;Systems and Control
|
1203.4553 | Isophote comprises a locus of the surface points whose normal vectors make a
constant angle with a fixed vector. Main objective of this paper is to find the
axis of an isophote curve via its Darboux frame and afterwards to give some
characterizations about isophote and its axis. Particularly, for isophotes
lying on a canal surface will be obtained other characterizations again.
| [
"math.DG"
] | math.DG | Differential Geometry | 2,010Differential Geometry
|
|
cs/9811020 | This paper addresses the issue of making legacy information (that material
held in paper format only) electronically searchable and retrievable. We used
proprietary software and commercial hardware to create a process for scanning,
cataloging, archiving and electronically disseminating full-text documents.
This process is relatively easy to implement and reasonably affordable.
| [
"cs.DL"
] | cs.DL | Digital Libraries | 2,081Digital Libraries
|
|
0910.3950 | The Constrained Minimal Supersymmetric Standard Model (CMSSM) is one of the
simplest and most widely-studied supersymmetric extensions to the standard
model of particle physics. Nevertheless, current data do not sufficiently
constrain the model parameters in a way completely independent of priors,
statistical measures and scanning techniques. We present a new technique for
scanning supersymmetric parameter spaces, optimised for frequentist profile
likelihood analyses and based on Genetic Algorithms. We apply this technique to
the CMSSM, taking into account existing collider and cosmological data in our
global fit. We compare our method to the MultiNest algorithm, an efficient
Bayesian technique, paying particular attention to the best-fit points and
implications for particle masses at the LHC and dark matter searches. Our
global best-fit point lies in the focus point region. We find many
high-likelihood points in both the stau co-annihilation and focus point
regions, including a previously neglected section of the co-annihilation region
at large m_0. We show that there are many high-likelihood points in the CMSSM
parameter space commonly missed by existing scanning techniques, especially at
high masses. This has a significant influence on the derived confidence regions
for parameters and observables, and can dramatically change the entire
statistical inference of such scans.
| [
"hep-ph",
"astro-ph.CO"
] | hep-ph | astro-ph.CO | High Energy Physics - Phenomenology;Cosmology and Nongalactic Astrophysics | 3,156High Energy Physics - Phenomenology;Cosmology and Nongalactic Astrophysics
|
2111.06718 | In this paper we present a short overview of the new Wolfram Mathematica
package intended for elementary "in-basis" tensor and differential-geometric
calculations. In contrast to alternatives our package is designed to be
easy-to-use, short, all-purpose, and hackable. It supports tensor contractions
using Einstein notation, transformations between different bases, tensor
derivative operator, expansion in basis vectors and forms, exterior derivative,
and interior product.
| [
"nucl-th",
"cs.MS",
"cs.SC",
"hep-th",
"physics.comp-ph"
] | nucl-th | cs.MS | Nuclear Theory;Mathematical Software;Symbolic Computation;High Energy Physics - Theory;Computational Physics | 7,267longtail
|
cond-mat/0403387 | An exact treatment of the non-equilibrium dynamics of hard-core bosons on
one-dimensional lattices shows that, starting from a pure Fock state,
quasi-long-range correlations develop dynamically, and that they lead to the
formation of quasi-condensates at finite momenta. Scaling relations
characterizing the quasi-condensate and the dynamics of its formation are
obtained. The relevance of our findings for atom lasers with full control of
the wave-length by means of a lattice is discussed.
| [
"cond-mat.stat-mech"
] | cond-mat.stat-mech | Statistical Mechanics | 6,821Statistical Mechanics
|
|
1611.01091 | In 1982, Tamaki Yano proposed a conjecture predicting the set of b-exponents
of an irreducible plane curve singularity germ which is generic in its
equisingularity class. In \cite{ACLM-Yano2} we proved the conjecture for the
case in which the germ has two Puiseux pairs and its algebraic monodromy has
distinct eigenvalues. In this article we aim to study the Bernstein polynomial
for any function with the hypotheses above. In particular the set of all common
roots of those Bernstein polynomials is given. We provide also bounds for some
analytic invariants of singularities and illustrate the computations in
suitable examples.
| [
"math.AG"
] | math.AG | Algebraic Geometry | 47Algebraic Geometry
|
|
1505.01483 | We present Atacama Large Millimeter/submillimeter Array (ALMA) observations
of GSC 6214-210 A and B, a solar-mass member of the 5-10 Myr Upper Scorpius
association with a 15 $\pm$ 2 Mjup companion orbiting at $\approx$330 AU
(2.2"). Previous photometry and spectroscopy spanning 0.3-5 $\mu$m revealed
optical and thermal excess as well as strong H$\alpha$ and Pa~$\beta$ emission
originating from a circum-substellar accretion disk around GSC 6214-210 B,
making it the lowest mass companion with unambiguous evidence of a subdisk.
Despite ALMA's unprecedented sensitivity and angular resolution, neither
component was detected in our 880 $\mu$m (341 GHz) continuum observations down
to a 3-$\sigma$ limit of 0.22 mJy/beam. The corresponding constraints on the
dust mass and total mass are <0.15 Mearth and <0.05 Mjup, respectively, or
<0.003% and <0.3% of the mass of GSC 6214-210 B itself assuming a 100:1
gas-to-dust ratio and characteristic dust temperature of 10-20 K. If the host
star possesses a putative circum-stellar disk then at most it is a meager
0.0015% of the primary mass, implying that giant planet formation has certainly
ceased in this system. Considering these limits and its current accretion rate,
GSC 6214-210 B appears to be at the end stages of assembly and is not expected
to gain any appreciable mass over the next few Myr.
| [
"astro-ph.EP"
] | astro-ph.EP | Earth and Planetary Astrophysics | 2,351Earth and Planetary Astrophysics
|
|
1912.08868 | We describe the use of Non-Negative Matrix Factorization (NMF) and Latent
Dirichlet Allocation (LDA) algorithms to perform topic mining and labelling
applied to retail customer communications in attempt to characterize the
subject of customers inquiries. In this paper we compare both algorithms in the
topic mining performance and propose methods to assign topic subject labels in
an automated way.
| [
"cs.LG",
"cs.CL",
"cs.CY",
"cs.IR",
"stat.ML"
] | cs.LG | cs.CL | Machine Learning;Computation and Language;Computers and Society;Information Retrieval;Machine Learning | 7,267longtail
|
1202.0388 | We present colour transformations for the conversion of the Wide-Field Survey
Explorer (WISE) W1, W2, and W3 magnitudes to the Johnson-Cousins (BVIc), Sloan
Digital Sky Survey (gri), and Two Micron All Sky Survey JHKs photometric
systems, for red clump (RC) stars. RC stars were selected from the Third Radial
Velocity Experiment (RAVE) Data Release (DR3). The apparent magnitudes were
collected by matching the coordinates of this sample with different photometric
catalogues. The final sample (355 RC stars) used to obtain metallicity
dependent-and free of metallicity- transformations between WISE and
Johnson-Cousins, SDSS, 2MASS photometric systems. These transformations
combined with known absolute magnitudes at shorter wavelengths can be used in
space density determinations for the Galactic (thin and thick) discs at
distances larger than the ones evaluated with JHKs photometry alone, hence
providing a powerful tool in the analysis of Galactic structure.
| [
"astro-ph.GA"
] | astro-ph.GA | Astrophysics of Galaxies | 464Astrophysics of Galaxies
|
|
2207.13769 | This paper presents the results of calculating Casimir-Lifshitz friction
force and heating rate of a small metal particle moving above a metal surface
(thick plate) in the case of their different local temperatures. The case of
normal nonmagnetic metals (Au) is considered. There is a strong interplay of
temperatures, particle velocity and separation distance resulting in an
anomalous direction of the heat flux between the bodies and a peak temperature
dependence of the friction force at sufficiently low temperatures. In
particular, a hot moving particle can additionally receive heat from a cold
surface. The conditions for experimental measurement of these effects are
discussed.
| [
"cond-mat.mes-hall"
] | cond-mat.mes-hall | Mesoscale and Nanoscale Physics | 4,450Mesoscale and Nanoscale Physics
|
|
2310.20616 | Cross-flow turbines harness kinetic energy in wind or moving water. Due to
their unsteady fluid dynamics, it can be difficult to predict the interplay
between aspects of rotor geometry and turbine performance. This study considers
the effects of three geometric parameters: the number of blades, the preset
pitch angle, and the chord-to-radius ratio. The relevant fluid dynamics of
cross-flow turbines are reviewed, as are prior experimental studies that have
investigated these parameters in a more limited manner. Here, 223 unique
experiments are conducted across an order of magnitude of diameter-based
Reynolds numbers ($\approx 8\!\times\!10^4 - 8\!\times\!10^5$) in which the
performance implications of these three geometric parameters are evaluated. In
agreement with prior work, maximum performance is generally observed to
increase with Reynolds number and decrease with blade count. The broader
experimental space identifies new parametric interdependencies; for example,
the optimal preset pitch angle is increasingly negative as the chord-to-radius
ratio increases. Because these experiments vary both the chord-to-radius ratio
and blade count, the performance of different rotor geometries with the same
solidity (the ratio of total blade chord to rotor circumference) can also be
evaluated. Results demonstrate that while solidity can be a poor predictor of
maximum performance, across all scales and tested geometries it is an excellent
predictor of the tip-speed ratio corresponding to maximum performance. Overall,
these results present a uniquely holistic view of relevant geometric
considerations for cross-flow turbine rotor design and provide a rich dataset
for validation of numerical simulations and reduced-order models.
| [
"physics.flu-dyn"
] | physics.flu-dyn | Fluid Dynamics | 2,452Fluid Dynamics
|
|
2201.08416 | We study the electromechanical response of Janus transition metal
dichalcogenide (TMD) nanotubes from first principles. In particular,
considering both armchair and zigzag variants of twenty-seven select Janus TMD
nanotubes, we determine the change in bandgap and charge carriers' effective
mass upon axial and torsional deformations using density functional theory
(DFT). We observe that metallic nanotubes remain unaffected, whereas the
bandgap in semiconducting nanotubes decreases linearly and quadratically with
axial and shear strains, respectively, leading to semiconductor--metal
transitions. In addition, we find that there is a continuous decrease and
increase in the effective mass of holes and electrons with strains,
respectively, leading to n-type--p-type semiconductor transitions. We show that
this behavior is a consequence of the rehybridization of orbitals, rather than
charge transfer between the atoms. Overall, mechanical deformations form a
powerful tool for tailoring the electronic response of semiconducting Janus TMD
nanotubes.
| [
"cond-mat.mtrl-sci",
"physics.chem-ph"
] | cond-mat.mtrl-sci | physics.chem-ph | Materials Science;Chemical Physics | 4,301Materials Science;Chemical Physics
|
1712.00819 | We establish a theoretical framework for exploring the quantum dynamics of
finite ultracold bosonic ensembles based on the
Born-Bogoliubov-Green-Kirkwood-Yvon (BBGKY) hierarchy of equations of motion
for few-particle reduced density matrices (RDMs). The theory applies to zero as
well as low temperatures and is formulated in a highly efficient way by
utilizing dynamically optimized single-particle basis states and representing
the RDMs in terms of permanents with respect to those. An energy, RDM
compatibility and symmetry conserving closure approximation is developed on the
basis of a recursively formulated cluster expansion for these finite systems.
In order to enforce necessary representability conditions, two novel,
minimal-invasive and energy-conserving correction algorithms are proposed,
involving the dynamical purification of the solution of the truncated BBGKY
hierarchy and the correction of the equations of motion themselves,
respectively. For gaining conceptual insights, the impact of two-particle
correlations on the dynamical quantum depletion is studied analytically. We
apply this theoretical framework to both a tunneling and an interaction-quench
scenario. Due to our efficient formulation of the theory, we can reach
truncation orders as large as twelve and thereby systematically study the
impact of the truncation order on the results. While the short-time dynamics is
found to be excellently described with controllable accuracy, significant
deviations occur on a longer time-scale in sufficiently far off-equilibrium
situations. Theses deviations are accompanied by exponential-like instabilities
leading to unphysical results. The phenomenology of these instabilities is
investigated in detail and we show that the minimal-invasive correction
algorithm of the equation of motion can indeed stabilize the BBGKY hierarchy
truncated at the second order.
| [
"quant-ph",
"cond-mat.quant-gas"
] | quant-ph | cond-mat.quant-gas | Quantum Physics;Quantum Gases | 6,169Quantum Physics;Quantum Gases
|
2311.00790 | Metaphor identification aims at understanding whether a given expression is
used figuratively in context. However, in this paper we show how existing
metaphor identification datasets can be gamed by fully ignoring the potential
metaphorical expression or the context in which it occurs. We test this
hypothesis in a variety of datasets and settings, and show that metaphor
identification systems based on language models without complete information
can be competitive with those using the full context. This is due to the
construction procedures to build such datasets, which introduce unwanted biases
for positive and negative classes. Finally, we test the same hypothesis on
datasets that are carefully sampled from natural corpora and where this bias is
not present, making these datasets more challenging and reliable.
| [
"cs.CL"
] | cs.CL | Computation and Language | 1,168Computation and Language
|
|
1708.06602 | This paper investigates the energy bounds in modified Gauss-Bonnet gravity
with anisotropic background. Locally rotationally symmetric Bianchi type ${I}$
cosmological model in $f(R,G)$ gravity is considered to meet this aim.
Primarily, a general $f(R,G)$ model is used to develop the field equations. In
this aspect, we investigate the viability of modified gravitational theory by
studying the energy conditions. We take in account four $f(R,G)$ gravity models
commonly discussed in the literature. We formulate the inequalities obtained by
energy conditions and investigate the viability of the above mentioned models
using the Hubble, deceleration, jerk and snap parameters. Graphical analysis
shows that for first two $f(R,G)$ gravity models, NEC, WEC and SEC are
satisfied under suitable values of anisotropy and model parameters involved.
Moreover, SEC is violated for the third and fourth models which predicts the
cosmic expansion.
| [
"physics.gen-ph"
] | physics.gen-ph | General Physics | 2,645General Physics
|
|
2009.05138 | In many online platforms, customers' decisions are substantially influenced
by product rankings as most customers only examine a few top-ranked products.
Concurrently, such platforms also use the same data corresponding to customers'
actions to learn how these products must be ranked or ordered. These
interactions in the underlying learning process, however, may incentivize
sellers to artificially inflate their position by employing fake users, as
exemplified by the emergence of click farms. Motivated by such fraudulent
behavior, we study the ranking problem of a platform that faces a mixture of
real and fake users who are indistinguishable from one another. We first show
that existing learning algorithms---that are optimal in the absence of fake
users---may converge to highly sub-optimal rankings under manipulation by fake
users. To overcome this deficiency, we develop efficient learning algorithms
under two informational environments: in the first setting, the platform is
aware of the number of fake users, and in the second setting, it is agnostic to
the number of fake users. For both these environments, we prove that our
algorithms converge to the optimal ranking, while being robust to the
aforementioned fraudulent behavior; we also present worst-case performance
guarantees for our methods, and show that they significantly outperform
existing algorithms. At a high level, our work employs several novel approaches
to guarantee robustness such as: (i) constructing product-ordering graphs that
encode the pairwise relationships between products inferred from the customers'
actions; and (ii) implementing multiple levels of learning with a judicious
amount of bi-directional cross-learning between levels.
| [
"cs.LG",
"cs.IR",
"stat.ML"
] | cs.LG | cs.IR | Machine Learning;Information Retrieval;Machine Learning | 4,156Machine Learning;Information Retrieval;Machine Learning
|
2007.14328 | We present an option pricing formula for European options in a stochastic
volatility model. In particular, the volatility process is defined using a
fractional integral of a diffusion process and both the stock price and the
volatility processes have jumps in order to capture the market effect known as
leverage effect. We show how to compute a martingale representation for the
volatility process. Finally, using It\^o calculus for processes with
discontinuous trajectories, we develop a first order approximation formula for
option prices. There are two main advantages in the usage of such approximating
formulas to traditional pricing methods. First, to improve computational
effciency, and second, to have a deeper understanding of the option price
changes in terms of changes in the model parameters.
| [
"q-fin.PR",
"math.PR"
] | q-fin.PR | math.PR | Pricing of Securities;Probability | 5,704Pricing of Securities;Probability
|
2101.09155 | In this paper we derive some Edmundson-Lah-Ribari\v{c} type inequalities for
positive linear functionals and 3-convex functions. Main results are applied to
the generalized f-divergence functional. Examples with Zipf Mandelbrot law are
used to illustrate the results. In addition, obtained results are utilized in
constructing some families of exponentially convex functions and Stolarsky-type
means.
| [
"math.GM"
] | math.GM | General Mathematics | 2,639General Mathematics
|
|
1809.10196 | Objective: Ultrahigh-resolution optical coherence microscopy (OCM) has
recently demonstrated its potential for accurate diagnosis of human cervical
diseases. One major challenge for clinical adoption, however, is the steep
learning curve clinicians need to overcome to interpret OCM images. Developing
an intelligent technique for computer-aided diagnosis (CADx) to accurately
interpret OCM images will facilitate clinical adoption of the technology and
improve patient care. Methods: 497 high-resolution 3-D OCM volumes (600
cross-sectional images each) were collected from 159 ex vivo specimens of 92
female patients. OCM image features were extracted using a convolutional neural
network (CNN) model, concatenated with patient information (e.g., age, HPV
results), and classified using a support vector machine classifier. Ten-fold
cross-validations were utilized to test the performance of the CADx method in a
five-class classification task and a binary classification task. Results: An
88.3 plus or minus 4.9% classification accuracy was achieved for five
fine-grained classes of cervical tissue, namely normal, ectropion, low-grade
and high-grade squamous intraepithelial lesions (LSIL and HSIL), and cancer. In
the binary classification task (low-risk [normal, ectropion and LSIL] vs.
high-risk [HSIL and cancer]), the CADx method achieved an area-under-the-curve
(AUC) value of 0.959 with an 86.7 plus or minus 11.4% sensitivity and 93.5 plus
or minus 3.8% specificity. Conclusion: The proposed deep-learning based CADx
method outperformed three human experts. It was also able to identify
morphological characteristics in OCM images that were consistent with
histopathological interpretations. Significance: Label-free OCM imaging,
combined with deep-learning based CADx methods, hold a great promise to be used
in clinical settings for the effective screening and diagnosis of cervical
diseases.
| [
"cs.CV"
] | cs.CV | Computer Vision and Pattern Recognition | 1,498Computer Vision and Pattern Recognition
|
|
1510.00109 | We consider a multi-agent confinement control problem in which a single
leader has a purely repulsive effect on follower agents with double-integrator
dynamics. By decomposing the leader's control inputs into periodic and
aperiodic components, we show that the leader can be driven so as to guarantee
confinement of the followers about a time-dependent trajectory in the plane. We
use tools from averaging theory and an input-to-state stability type argument
to derive conditions on the model parameters that guarantee confinement of the
followers about the trajectory. For the case of a single follower, we show that
if the follower starts at the origin, then the error in trajectory tracking can
be made arbitrarily small depending on the frequency of the periodic control
components and the rate of change of the trajectory. We validate our approach
using simulations and experiments with a small mobile robot.
| [
"cs.MA",
"cs.SY",
"math.OC"
] | cs.MA | cs.SY | Multiagent Systems;Systems and Control;Optimization and Control | 4,690Multiagent Systems;Systems and Control;Optimization and Control
|
chao-dyn/9907032 | Separated flows past complex geometries are modelled by discrete vortex
techniques. The flows are assumed to be rotational and inviscid, and a new
technique is described to determine the streamfunctions for linear shear
profiles. The geometries considered are the snow cornice and the
backward-facing step, whose edges allow for the separation of the flow and
reattachment downstream of the recirculation regions. A point vortex has been
added to the flows in order to constrain the separation points to be located at
the edges, while the conformal mappings have been modified in order to smooth
the sharp edges and let the separation points be free to oscillate around the
points of maximum curvature. Unsteadiness is imposed on the flow by perturbing
the vortex location, either by displacing the vortex from equilibrium, or by
imposing a random perturbation with zero mean on the vortex in equilibrium. The
trajectories of passive scalars continuously released upwind of the separation
point and trapped by the recirculating bubble are numerically integrated, and
concentration time series are calculated at fixed locations downwind of the
reattachment points. This model proves to be capable of reproducing the
trapping and intermittent release of scalars, in agreement with the simulation
of the flow past a snow cornice performed by a discrete multi-vortex model, as
well as with direct numerical simulations of the flow past a backward-facing
step. The simulation results indicate that for flows undergoing separation and
reattachment the unsteadiness of the recirculating bubble is the main mechanism
responsible for the intense large-scale concentration fluctuations downstream.
| [
"chao-dyn",
"nlin.CD"
] | chao-dyn | nlin.CD | Chaotic Dynamics;Chaotic Dynamics | 821Chaotic Dynamics;Chaotic Dynamics
|
cond-mat/0107264 | We use the two-electron wavefunctions (geminals) and the simple screened
Coulomb potential proposed by Overhauser [Can. J. Phys. 73, 683 (1995)] to
compute the pair-distribution function for a uniform electron gas. We find
excellent agreement with Quantum Monte Carlo simulations in the short-range
region, for a wide range of electron densities. We are thus able to estimate
the value of the second-order coefficient of the small interelectronic-distance
expansion of the pair-distribution function. The results are generalized to the
partially polarized gas.
| [
"cond-mat"
] | cond-mat | Condensed Matter | 1,697Condensed Matter
|
|
1602.00680 | We perform the one-loop induced charged lepton flavor violating decays of the
neutral Higgses in an extended mirror fermion model with non-sterile
electroweak-scale right-handed neutrinos and a horizontal $A_4$ symmetry in the
lepton sector. We demonstrate that for the 125 GeV scalar $h$ there is tension
between the recent LHC result ${\cal B}(h \to \tau \mu) \sim $ 1% and the
stringent limits on the rare processes $\mu \to e \gamma$ and $\tau \to (\mu$
or $e) \gamma$ from low energy experiments.
| [
"hep-ph"
] | hep-ph | High Energy Physics - Phenomenology | 3,129High Energy Physics - Phenomenology
|
|
1906.08916 | We present a melody based classification of musical styles by exploiting the
pitch and energy based characteristics derived from the audio signal. Three
prominent musical styles were chosen which have improvisation as integral part
with similar melodic principles, theme, and structure of concerts namely,
Hindustani, Carnatic and Turkish music. Listeners of one or more of these
genres can discriminate between these based on the melodic contour alone.
Listening tests were carried out using melodic attributes alone, on similar
melodic pieces with respect to raga/makam, and removing any instrumentation cue
to validate our hypothesis that style distinction is evident in the melody. Our
method is based on finding a set of highly discriminatory features, derived
from musicology, to capture distinct characteristics of the melodic contour.
Behavior in terms of transitions of the pitch contour, the presence of
micro-tonal notes and the nature of variations in the vocal energy are
exploited. The automatically classified style labels are found to correlate
well with subjective listening judgments. This was verified by using
statistical tests to compare the labels from subjective and objective
judgments. The melody based features, when combined with timbre based features,
were seen to improve the classification performance.
| [
"cs.SD",
"eess.AS"
] | cs.SD | eess.AS | Sound;Audio and Speech Processing | 6,734Sound;Audio and Speech Processing
|
2204.06412 | We present a simplified way to access and manipulate the topology of massive
Dirac fermions by means of scalar potential. We show systematically how a
distribution of scalar potential can manipulate the signature of the gap or the
mass term as well as the dispersion leading to a band inversion via inverse
Klein tunnelling. In one dimension it can lead to the formation of edge
localisation. In two dimensions this can give rise to an emergent mechanism,
which we refer to as the Scalar Hall Effect. This can facilitate a direct
manipulation of topological invariants, e.g. the Chern number, as well as
allows to manipulate the edge states locally and thus opens new possibilities
for tuning physical observables which originate from the nontrivial topology.
| [
"cond-mat.mes-hall"
] | cond-mat.mes-hall | Mesoscale and Nanoscale Physics | 4,450Mesoscale and Nanoscale Physics
|
|
1509.03141 | Given a Banach space~$X$ with an unconditional basis, we consider the
following question: does the identity on~$X$ factor through every operator
on~$X$ with large diagonal relative to the unconditional basis? We show that on
Gowers' unconditional Banach space, there exists an operator for which the
answer to the question is negative. By contrast, for any operator on the
mixed-norm Hardy spaces $H^p(H^q)$, where $1 \leq p,q < \infty$, with the
bi-parameter Haar system, this problem always has a positive solution. The
spaces $L^p, 1 < p < \infty$, were treated first by Andrew~[{\em Studia
Math.}~1979].
| [
"math.FA"
] | math.FA | Functional Analysis | 2,549Functional Analysis
|
|
1711.11039 | The fundamental metallicity relation (FMR) is a postulated correlation
between galaxy stellar mass, star formation rate (SFR), and gas-phase
metallicity. At its core, this relation posits that offsets from the
mass-metallicity relation (MZR) at a fixed stellar mass are correlated with
galactic SFR. In this Letter, we quantify the timescale with which galactic
SFRs and metallicities evolve using hydrodynamical simulations. We find that
Illustris and IllustrisTNG predict that galaxy offsets from the star formation
main sequence and MZR evolve over similar timescales, are often anti-correlated
in their evolution, evolve with the halo dynamical time, and produce a
pronounced FMR. In fact, for a FMR to exist, the metallicity and SFR must
evolve in an anti-correlated sense which requires that they evolve with similar
time variability. In contrast to Illustris and IllustrisTNG, we speculate that
the SFR and metallicity evolution tracks may become decoupled in galaxy
formation models dominated by globally-bursty SFR histories, which could weaken
the FMR residual correlation strength. This opens the possibility of
discriminating between bursty and non-bursty feedback models based on the
strength and persistence of the FMR -- especially at high redshift.
| [
"astro-ph.GA"
] | astro-ph.GA | Astrophysics of Galaxies | 464Astrophysics of Galaxies
|
|
2207.09061 | Feature selection is an important process in machine learning. It builds an
interpretable and robust model by selecting the features that contribute the
most to the prediction target. However, most mature feature selection
algorithms, including supervised and semi-supervised, fail to fully exploit the
complex potential structure between features. We believe that these structures
are very important for the feature selection process, especially when labels
are lacking and data is noisy.
To this end, we innovatively introduce a deep learning-based self-supervised
mechanism into feature selection problems, namely batch-Attention-based
Self-supervision Feature Selection(A-SFS). Firstly, a multi-task
self-supervised autoencoder is designed to uncover the hidden structure among
features with the support of two pretext tasks. Guided by the integrated
information from the multi-self-supervised learning model, a batch-attention
mechanism is designed to generate feature weights according to batch-based
feature selection patterns to alleviate the impacts introduced by a handful of
noisy data. This method is compared to 14 major strong benchmarks, including
LightGBM and XGBoost. Experimental results show that A-SFS achieves the highest
accuracy in most datasets. Furthermore, this design significantly reduces the
reliance on labels, with only 1/10 labeled data needed to achieve the same
performance as those state of art baselines. Results show that A-SFS is also
most robust to the noisy and missing data.
| [
"cs.LG",
"cs.AI"
] | cs.LG | cs.AI | Machine Learning;Artificial Intelligence | 3,892Machine Learning;Artificial Intelligence
|
1701.08053 | Performance evaluation is a key issue for designers and users of Database
Management Systems (DBMSs). Performance is generally assessed with software
benchmarks that help, e.g., test architectural choices, compare different
technologies or tune a system. In the particular context of data warehousing
and On-Line Analytical Processing (OLAP), although the Transaction Processing
Performance Council (TPC) aims at issuing standard decision-support benchmarks,
few benchmarks do actually exist. We present in this chapter the Data Warehouse
Engineering Benchmark (DWEB), which allows generating various ad-hoc synthetic
data warehouses and workloads. DWEB is fully parameterized to fulfill various
data warehouse design needs. However, two levels of parameterization keep it
relatively easy to tune. We also expand on our previous work on DWEB by
presenting its new Extract, Transform, and Load (ETL) feature as well as its
new execution protocol. A Java implementation of DWEB is freely available
on-line, which can be interfaced with most existing relational DMBSs. To the
best of our knowledge, DWEB is the only easily available, up-to-date benchmark
for data warehouses.
| [
"cs.DB"
] | cs.DB | Databases | 1,977Databases
|
|
1912.03635 | This paper deals with study of Birkhoff-James orthogonality of a linear
operator to a subspace of operators defined between arbitrary Banach spaces. In
case the domain space is reflexive and the subspace is finite dimensional we
obtain a complete characterization. For arbitrary Banach spaces, we obtain the
same under some additional conditions. For arbitrary Hilbert space $
\mathbb{H},$ we also study orthogonality to subspace of the space of linear
operators $L(\mathbb{H}), $ both with respect to operator norm as well as
numerical radius norm.
| [
"math.FA"
] | math.FA | Functional Analysis | 2,549Functional Analysis
|
|
0907.5604 | In this paper, we give the first detailed proof of the short-time existence
of Deane Yang's local Ricci flow. Then using the local Ricci flow, we prove
short-time existence of the Ricci flow on noncompact manifolds, whose Ricci
curvature has global lower bound and sectional curvature has only local average
integral bound. The short-time existence of the Ricci flow on noncompact
manifolds with bounded curvature was studied by Wan-Xiong Shi in 1990s. As a
corollary of our main theorem, we get the short-time existence part of Shi's
theorem in this more general context.
| [
"math.DG",
"math.AP"
] | math.DG | math.AP | Differential Geometry;Analysis of PDEs | 2,022Differential Geometry;Analysis of PDEs
|
2307.01207 | Recommender Systems (RS) currently represent a fundamental tool in online
services, especially with the advent of Online Social Networks (OSN). In this
case, users generate huge amounts of contents and they can be quickly
overloaded by useless information. At the same time, social media represent an
important source of information to characterize contents and users' interests.
RS can exploit this information to further personalize suggestions and improve
the recommendation process. In this paper we present a survey of Recommender
Systems designed and implemented for Online and Mobile Social Networks,
highlighting how the use of social context information improves the
recommendation task, and how standard algorithms must be enhanced and optimized
to run in a fully distributed environment, as opportunistic networks. We
describe advantages and drawbacks of these systems in terms of algorithms,
target domains, evaluation metrics and performance evaluations. Eventually, we
present some open research challenges in this area.
| [
"cs.IR",
"cs.LG",
"cs.SI"
] | cs.IR | cs.LG | Information Retrieval;Machine Learning;Social and Information Networks | 3,614Information Retrieval;Machine Learning;Social and Information Networks
|
1410.1616 | In this paper we study the supersymmetric generalization of the new soft
theorem which was proposed by Cachazo and Strominger recently. At tree level,
we prove the validity of the super soft theorems in both ${\cal N}=4$
super-Yang-Mills theory and ${\cal N}=8$ supergravity using super-BCFW
recursion relations. We verify these theorems exactly by showing some examples.
| [
"hep-th"
] | hep-th | High Energy Physics - Theory | 3,266High Energy Physics - Theory
|
|
0912.4667 | We study the substructure statistics of a representative sample of galaxy
clusters by means of two currently popular substructure characterisation
methods, power ratios and centroid shifts. We use the 31 clusters from the
REXCESS sample, compiled from the southern ROSAT All-Sky cluster survey REFLEX
with a morphologically unbiased selection in X-ray luminosity and redshift, all
of which have been reobserved with XMM-Newton. We investigate the uncertainties
of the substructure parameters and examine the dependence of the results on
projection effects, finding that the uncertainties of the parameters can be
quite substantial. Thus while the quantification of the dynamical state of
individual clusters with these parameters should be treated with extreme
caution, these substructure measures provide powerful statistical tools to
characterise trends of properties in large cluster samples. The centre shift
parameter, w, is found to be more sensitive in general. For the REXCESS sample
neither the occurence of substructure nor the presence of cool cores depends on
cluster mass. There is a significant anti-correlation between the existence of
substantial substructure and cool cores. The simulated clusters show on average
larger substructure parameters than the observed clusters, a trend that is
traced to the fact that cool regions are more pronounced in the simulated
clusters, leading to stronger substructure measures in merging clusters and
clusters with offset cores. Moreover, the frequency of cool regions is higher
in the simulations than in the observations, implying that the description of
the physical processes shaping cluster formation in the simulations requires
further improvement.
| [
"astro-ph.CO",
"astro-ph.HE"
] | astro-ph.CO | astro-ph.HE | Cosmology and Nongalactic Astrophysics;High Energy Astrophysical Phenomena | 1,749Cosmology and Nongalactic Astrophysics;High Energy Astrophysical Phenomena
|
2208.00805 | The paper investigates a dial-a-ride problem focusing on the residents of
large cities. These individuals have the opportunity to use a wide variety of
transportation modes. Because of this, ridepooling providers have to solve the
tradeoff between a high pooling rate and a small detour for customers to be
competitive. We provide a Branch-and-Cut algorithm for this problem setting and
introduce a new technique using information about already fixed paths to
identify infeasible solutions ahead of time and to improve lower bounds on the
arrival times at customer locations. By this, we are able to introduce
additional valid inequalities to improve the search. We evaluate our procedure
in an extensive computational study with up to 120 customers and ten vehicles.
Our procedure finds significantly more optimal solutions and better lower and
upper bounds in comparison with a mixed-integer programming formulation.
| [
"math.OC"
] | math.OC | Optimization and Control | 5,234Optimization and Control
|
|
1510.06083 | Variable selection is a fundamental task in statistical data analysis.
Sparsity-inducing regularization methods are a popular class of methods that
simultaneously perform variable selection and model estimation. The central
problem is a quadratic optimization problem with an l0-norm penalty. Exactly
enforcing the l0-norm penalty is computationally intractable for larger scale
problems, so dif- ferent sparsity-inducing penalty functions that approximate
the l0-norm have been introduced. In this paper, we show that viewing the
problem from a convex relaxation perspective offers new insights. In
particular, we show that a popular sparsity-inducing concave penalty function
known as the Minimax Concave Penalty (MCP), and the reverse Huber penalty
derived in a recent work by Pilanci, Wainwright and Ghaoui, can both be derived
as special cases of a lifted convex relaxation called the perspective
relaxation. The optimal perspective relaxation is a related minimax problem
that balances the overall convexity and tightness of approximation to the l0
norm. We show it can be solved by a semidefinite relaxation. Moreover, a
probabilistic interpretation of the semidefinite relaxation reveals connections
with the boolean quadric polytope in combinatorial optimization. Finally by
reformulating the l0-norm pe- nalized problem as a two-level problem, with the
inner level being a Max-Cut problem, our proposed semidefinite relaxation can
be realized by replacing the inner level problem with its semidefinite
relaxation studied by Goemans and Williamson. This interpretation suggests
using the Goemans-Williamson rounding procedure to find approximate solutions
to the l0-norm penalized problem. Numerical experiments demonstrate the
tightness of our proposed semidefinite relaxation, and the effectiveness of
finding approximate solutions by Goemans-Williamson rounding.
| [
"cs.LG",
"math.NA",
"math.OC",
"stat.ML"
] | cs.LG | math.NA | Machine Learning;Numerical Analysis;Optimization and Control;Machine Learning | 7,267longtail
|
2004.01553 | In this paper, we investigate the almost surely pointwise convergence problem
of free KdV equation, free wave equation, free elliptic and non-elliptic
Schr\"odinger equation respectively. We firstly establish some estimates
related to the Wiener decomposition of frequency spaces which are just Lemmas
2.1-2.6 in this paper. Secondly, by using Lemmas 2.1-2.6, 3.1, we establish the
probabilistic estimates of some random series which are just Lemmas 3.2-3.11 in
this paper. Finally, combining the density theorem in L$^{2}$ with Lemmas
3.2-3.11, we obtain almost surely pointwise convergence of the solutions to
corresponding equations with randomized initial data in $L^{2}$, which require
much less regularity of the initial data than the rough data case. At the same
time, we present the probabilistic density theorem, which is Lemma 3.11 in this
paper.
| [
"math.AP"
] | math.AP | Analysis of PDEs | 205Analysis of PDEs
|
|
2002.09908 | Let $n_0$ be 1 or 3. If a multiplicative function $f$ satisfies $f(p+q-n_0) =
f(p)+f(q)-f(n_0)$ for all primes $p$ and $q$, then $f$ is the identity function
$f(n)=n$ or a constant function $f(n)=1$.
| [
"math.NT"
] | math.NT | Number Theory | 4,945Number Theory
|
|
2202.00911 | To leverage the power of big data from source tasks and overcome the scarcity
of the target task samples, representation learning based on multi-task
pretraining has become a standard approach in many applications. However, up
until now, choosing which source tasks to include in the multi-task learning
has been more art than science. In this paper, we give the first formal study
on resource task sampling by leveraging the techniques from active learning. We
propose an algorithm that iteratively estimates the relevance of each source
task to the target task and samples from each source task based on the
estimated relevance. Theoretically, we show that for the linear representation
class, to achieve the same error rate, our algorithm can save up to a
\textit{number of source tasks} factor in the source task sample complexity,
compared with the naive uniform sampling from all source tasks. We also provide
experiments on real-world computer vision datasets to illustrate the
effectiveness of our proposed method on both linear and convolutional neural
network representation classes. We believe our paper serves as an important
initial step to bring techniques from active learning to representation
learning.
| [
"cs.LG",
"cs.AI"
] | cs.LG | cs.AI | Machine Learning;Artificial Intelligence | 3,892Machine Learning;Artificial Intelligence
|
hep-ph/9608325 | In this talk the Higgs boson effects in electroweak precision observables are
reviewed and the possibility of indirect information on the Higgs mass from
electroweak radiative corrections and precision data is discussed.
| [
"hep-ph"
] | hep-ph | High Energy Physics - Phenomenology | 3,129High Energy Physics - Phenomenology
|
|
2004.07373 | This article explores the parallels between improvisational theater (Improv)
and teaching in an Active Learning environment. It presents the notions of
Active Teaching as a natural complement to Active Learning, and discusses how
unexpected situations give rise to valuable Teaching Moments. These Teaching
Moments can be strategically utilized following the rules of Improv. This
article presents some examples of this in the Mathematics classroom, as well as
the implementation of an Improv Seminar in the Mathematics Department at the
University of California, Santa Cruz.
| [
"math.HO"
] | math.HO | History and Overview | 3,426History and Overview
|
|
2211.16786 | The image recapture attack is an effective image manipulation method to erase
certain forensic traces, and when targeting on personal document images, it
poses a great threat to the security of e-commerce and other web applications.
Considering the current learning-based methods suffer from serious overfitting
problem, in this paper, we propose a novel two-branch deep neural network by
mining better generalized recapture artifacts with a designed frequency filter
bank and multi-scale cross-attention fusion module. In the extensive
experiment, we show that our method can achieve better generalization
capability compared with state-of-the-art techniques on different scenarios.
| [
"cs.CV",
"cs.AI"
] | cs.CV | cs.AI | Computer Vision and Pattern Recognition;Artificial Intelligence | 1,502Computer Vision and Pattern Recognition;Artificial Intelligence
|