May 2022 arXiv papers — page 26

Liam Hebert, Lukasz Golab, Pascal Poupart, Robin Cohen

A core issue in multi-agent federated reinforcement learning is defining how to aggregate insights from multiple agents. This is commonly done by taking the average of each participating agent's model weights into one common model (FedAvg). We instead propose FedFormer, a novel federation strategy that utilizes Transformer Attention to contextually aggregate

Zelin Zhang, Songbai Chen, Jiliang Jing

We have studied the image of Bonnor black dihole surrounded by a thin accretion disk where the electromagnetic emission is assumed to be dominated respectively by black body radiation and synchrotron radiation. Our results show that the intensity of Bonnor black dihole image increases with the magnetic parameter and the inclination angle in both radiation mo

Phuc D. Nguyen, Kristy L. Hansen, Branko Zajamsek, Peter Catcheside

One of the major sources of uncertainty in predictions of wind farm noise (WFN) reflect parametric and model structure uncertainty. The model structure uncertainty is a systematic uncertainty, which relates to uncertainty about the appropriate mathematical structure of the models. Here we quantified the model structure uncertainty in predicting WFN arising f

Xinze Li, Bruno Staffa

We show that given a closed $n$-manifold $M$, for a generic set of Riemannian metrics $g$ on $M$ there exists a sequence of closed geodesics that are equidistributed in $M$ if $n=2$; and an equidistributed sequence of embedded stationary geodesic nets if $n=3$. One of the main tools that we use is the Weyl Law for the volume spectrum for $1$-cycles, proved b

Yang Liu, Tian Liang, Yuxin Fu, Yu-Xin Xie

This paper proposes a new metamaterial structure consisting of a periodically porous elastomer with pore coatings. This design enables us to engender finite deformation by a contactless load. As a case study, we apply thermal load to the pore coating and carry out a finite element analysis to probe instabilities and the associated phononic properties. It tur

Liam Collins, Hamed Hassani, Aryan Mokhtari, Sanjay Shakkottai

The Federated Averaging (FedAvg) algorithm, which consists of alternating between a few local stochastic gradient updates at client nodes, followed by a model averaging update at the server, is perhaps the most commonly used method in Federated Learning. Notwithstanding its simplicity, several empirical studies have illustrated that the output model of FedAv

Nassim Derriche, Ilya Elfimov, George Sawatzky

The understanding of lattice instabilities is of vast importance in material science. The famous example is the Peierls instability of one-dimensional metals and for strongly-nested Fermi surfaces in two and three dimensions. Through an analysis of H and Li chains in band theory, we find that the Bloch wave nature of the wavefunctions, if involving strong k-

contribute equally, K. Tsukagoshi, T. Nabatame, Z. J. Ding

The Kikuchi bands arise from Bragg diffraction of incoherent electrons scattered within a crystalline specimen and can be observed in both the transmission and reflection modes of scanning electron microscopy (SEM). Converging, rocking, or grazing incidence beams must be used to generate divergent electron sources to obtain the Kikuchi pattern. This paper re

Subhojyoti Mukherjee

In this paper, we consider the setting of piecewise i.i.d. bandits under a safety constraint. In this piecewise i.i.d. setting, there exists a finite number of changepoints where the mean of some or all arms change simultaneously. We introduce the safety constraint studied in \citet{wu2016conservative} to this setting such that at any round the cumulative re

Kehui Li, David C. Spierings, Aephraim M. Steinberg

Adiabatic Rapid Passage (ARP) is a powerful technique for efficient transfer of population between quantum states. In the lab, the efficiency of ARP is often limited by noise on either the energies of the states or the frequency of the driving field. We study ARP in the simple setting of a two-level system subject to sinusoidal fluctuations on the energy lev

Sen Na, Michael W. Mahoney

We consider online statistical inference of constrained stochastic nonlinear optimization problems. We apply the Stochastic Sequential Quadratic Programming (StoSQP) method to solve these problems, which can be regarded as applying second-order Newton's method to the Karush-Kuhn-Tucker (KKT) conditions. In each iteration, the StoSQP method computes the Newto

Amit Sharma

The main objective of this paper is to show that the homotopy colimit of a diagram of quasi-categories and indexed by a small category is a localization of Lurie's higher Grothendieck construction of the diagram. We thereby generalize Thomason's classical result which states that the homotopy colimit of a diagram of categories has the homotopy type of (the c

Jiahe Lan, Rui Zhang, Zheng Yan, Jie Wang

Speaker recognition has become very popular in many application scenarios, such as smart homes and smart assistants, due to ease of use for remote control and economic-friendly features. The rapid development of SRSs is inseparable from the advancement of machine learning, especially neural networks. However, previous work has shown that machine learning mod

Carles Domingo-Enrich, Yair Schiff, Youssef Mroueh

Learning high-dimensional distributions is often done with explicit likelihood modeling or implicit modeling via minimizing integral probability metrics (IPMs). In this paper, we expand this learning paradigm to stochastic orders, namely, the convex or Choquet order between probability measures. Towards this end, exploiting the relation between convex orders

Katharina Burgdorf, Negar Rostamzadeh, Ramya Srinivasan, Jennifer Lena

How can researchers from the creative ML/AI community and sociology of culture engage in fruitful collaboration? How do researchers from both fields think (differently) about creativity and the production of creative work? While the ML community considers creativity as a matter of technical expertise and acumen, social scientists have emphasized the role of

Dmitry Petrov, Matheus Gadelha, Radomir Mech, Evangelos Kalogerakis

We present ANISE, a method that reconstructs a 3D~shape from partial observations (images or sparse point clouds) using a part-aware neural implicit shape representation. The shape is formulated as an assembly of neural implicit functions, each representing a different part instance. In contrast to previous approaches, the prediction of this representation p

Kunal Garg, Mayank Baranwal

Distributed optimization has gained significant attention in recent years, primarily fueled by the availability of a large amount of data and privacy-preserving requirements. This paper presents a fixed-time convergent optimization algorithm for solving a potentially non-convex optimization problem using a first-order multi-agent system. Each agent in the ne

Ali Shirali

Datasets are often generated in a sequential manner, where the previous samples and intermediate decisions or interventions affect subsequent samples. This is especially prominent in cases where there are significant human-AI interactions, such as in recommender systems. To characterize the importance of this relationship across samples, we propose to use ad

Gilad Cohen, Raja Giryes

Member inference (MI) attacks aim to determine if a specific data sample was used to train a machine learning model. Thus, MI is a major privacy threat to models trained on private sensitive data, such as medical records. In MI attacks one may consider the black-box settings, where the model's parameters and activations are hidden from the adversary, or the

Liren Yu, Jiaming Xu, Xiaojun Lin

There is a growing interest in designing Graph Neural Networks (GNNs) for seeded graph matching, which aims to match two unlabeled graphs using only topological information and a small set of seed nodes. However, most previous GNNs for this task use a semi-supervised approach, which requires a large number of seeds and cannot learn knowledge that is transfer

Robyn Ritchie, Alon Harell, Phil Shreeves

Passing during power plays in hockey is a crucial component to move one's team closer to scoring a goal. With the use of women's ice hockey event and tracking data from the elimination round games during the 2022 Winter Olympics, we evaluate passing and assess players' risk-reward behaviours in these high intensity moments. We develop a model for probabilist

David Q. Aruquipa, Marc Casals

Field perturbations of a curved background spacetime generally propagate not only at the speed of light but also at all smaller velocities. This so-called $Hadamard\,tail$ contribution to wave propagation is relevant in various settings, from classical self-force calculations to communication between quantum particle detectors. One method for calculating thi

Kyle Hayes, Michael W. Fouts, Ali Baheri, David S. Mebane

A promising approach for scalable Gaussian processes (GPs) is the Karhunen-Lo\`eve (KL) decomposition, in which the GP kernel is represented by a set of basis functions which are the eigenfunctions of the kernel operator. Such decomposed kernels have the potential to be very fast, and do not depend on the selection of a reduced set of inducing points. Howeve

Andre Xian Ming Chang, Parth Khopkar, Bashar Romanous, Abhishek Chaurasia

The Streaming Engine (SE) is a Coarse-Grained Reconfigurable Array which provides programming flexibility and high-performance with energy efficiency. An application program to be executed on the SE is represented as a combination of Synchronous Data Flow (SDF) graphs, where every instruction is represented as a node. Each node needs to be mapped to the righ

Ehsan Variani, Ke Wu, Michael Riley, David Rybach

We introduce the Globally Normalized Autoregressive Transducer (GNAT) for addressing the label bias problem in streaming speech recognition. Our solution admits a tractable exact computation of the denominator for the sequence-level normalization. Through theoretical and empirical results, we demonstrate that by switching to a globally normalized model, the

Chiara Drolsbach, Nicolas Pröllochs

The spread of misinformation on social media is a pressing societal problem that platforms, policymakers, and researchers continue to grapple with. As a countermeasure, recent works have proposed to employ non-expert fact-checkers in the crowd to fact-check social media content. While experimental studies suggest that crowds might be able to accurately asses

Sara Bahaadini, Yunan Wu, Scott Coughlin, Michael Zevin

In this paper, leveraging the capabilities of neural networks for modeling the non-linearities that exist in the data, we propose several models that can project data into a low dimensional, discriminative, and smooth manifold. The proposed models can transfer knowledge from the domain of known classes to a new domain where the classes are unknown. A cluster

Zijie Li, Kazem Meidani, Amir Barati Farimani

Data-driven learning of partial differential equations' solution operators has recently emerged as a promising paradigm for approximating the underlying solutions. The solution operators are usually parameterized by deep learning models that are built upon problem-specific inductive biases. An example is a convolutional or a graph neural network that exploit

Shiau-Jie Rau, Kuo-Chuan Pan

Recent theoretical and numerical studies of Type Ia supernova explosion within the single-degenerate scenario suggest that the non-degenerate companions could survive during the supernova impact and could be detectable in nearby supernova remnants. However, observational efforts show less promising evidence on the existence of surviving companions from the s

William Matzko, Shobita Satyapal, Sara L. Ellison, Remington O. Sexton

Powerful outflows are thought to play a critical role in galaxy evolution and black hole growth. We present the first large-scale systematic study of ionised outflows in paired galaxies and post-mergers compared to a robust control sample of isolated galaxies. We isolate the impact of the merger environment to determine if outflow properties depend on merger

Xiaochen Hong, Mahdi Behnami, Long Yuan, Boqiang Li

The spin-1/2 kagom\'{e} Heisenberg antiferromagnet is generally accepted as one of the most promising two-dimensional models to realize a quantum spin liquid state. Previous experimental efforts were almost exclusively on only one archetypal material, the herbertsmithite ZnCu$_3$(OH)$_6$Cl$_2$, which unfortunately suffers from the notorious orphan spins prob

B. Flebus, A. H. MacDonald

When time-reversal symmetry is broken, the low-energy description of acoustic lattice dynamics allows for a dissipationless component of the viscosity tensor, the phonon Hall viscosity, which captures how phonon chirality grows with the wavevector. In this work, we show that, in ionic crystals, a phonon Hall viscosity contribution is produced by the Lorentz

Pablo Andújar Guerrero

Let $\mathcal{S}$ be a family of sets with VC-codensity less than $2$. We prove that, if $\mathcal{S}$ has the $(\omega, 2)$-property (for any infinitely many sets in $\mathcal{S}$, at least $2$ among them intersect), then $\mathcal{S}$ can be partitioned into finitely many subfamilies, each with the finite intersection property. If $\mathcal{S}$ is definabl

Vincent Dufour-Décieux, Christopher Moakler, Maria Cameron, Evan J. Reed

Hydrocarbon pyrolysis is a complex process involving large numbers of chemical species and types of chemical reactions. Its quantitative description is important for planetary sciences, in particular, for understanding the processes occurring in the interior of icy planets, such as Uranus and Neptune, where small hydrocarbons are subjected to high temperatur

Xiaochen Hong, Steffen Sykora, Federico Caglieris, Mahdi Behnami

Nematicity in the heavily hole-doped iron pnictide superconductors remains controversial. Sizeable nematic fluctuations and even nematic orders far from a magnetic instability were declared in RbFe$_2$As$_2$ and its sister compounds. Here we report a systematic elastoresistance study of series of isovalent- and electron-doped KFe$_2$As$_2$ crystals. We found

Robert Hu, Siu Lun Chau, Jaime Ferrando Huertas, Dino Sejdinovic

While preference modelling is becoming one of the pillars of machine learning, the problem of preference explanation remains challenging and underexplored. In this paper, we propose \textsc{Pref-SHAP}, a Shapley value-based model explanation framework for pairwise comparison data. We derive the appropriate value functions for preference models and further ex

Michal Macha, Sanjin Marion, Mukesh Tripathi, Mukeshchand Thakur

Large-area nanopore drilling is a major bottleneck in state-of-the-art nanoporous 2D membrane fabrication protocols. In addition, high-quality structural and statistical descriptions of as-fabricated porous membranes are key to predicting the corresponding membrane-wide permeation properties. In this work, we investigate Xe-ion focused ion beam as a tool for

Kanthashree Mysore Sathyendra, Thejaswi Muniyappa, Feng-Ju Chang, Jing Liu

Personal rare word recognition in end-to-end Automatic Speech Recognition (E2E ASR) models is a challenge due to the lack of training data. A standard way to address this issue is with shallow fusion methods at inference time. However, due to their dependence on external language models and the deterministic approach to weight boosting, their performance is

Hao Zhou, Yaozhong Hu, Yanghui Liu

We study the traditional backward Euler method for $m$-dimensional stochastic differential equations driven by fractional Brownian motion with Hurst parameter $H > 1/2$ whose drift coefficient satisfies the one-sided Lipschitz condition. The backward Euler scheme is proved to be of order $1$ and this rate is optimal by showing the asymptotic error distributi

Rediet Abebe, Nicole Immorlica, Jon Kleinberg, Brendan Lucier

The tendency for individuals to form social ties with others who are similar to themselves, known as homophily, is one of the most robust sociological principles. Since this phenomenon can lead to patterns of interactions that segregate people along different demographic dimensions, it can also lead to inequalities in access to information, resources, and op

Jose A. Arango-Sánchez, Julián D. Arias-Londoño

This work addresses the problem of speech separation in the Spanish Language using pre-trained deep learning models. As with many speech processing tasks, large databases in other languages different from English are scarce. Therefore this work explores different training strategies using the Conv-TasNet model as a benchmark. A scale-invariant signal distort

Félix del Teso, Erik Lindgren

We propose a new finite difference scheme for the degenerate parabolic equation \[ \partial_t u - \mbox{div}(|\nabla u|^{p-2}\nabla u) =f, \quad p\geq 2. \] Under the assumption that the data is H\"older continuous, we establish the convergence of the explicit-in-time scheme for the Cauchy problem provided a suitable stability type CFL-condition. An importan

M. De Sanctis

A reduced form of the Dirac equation has been previously introduced and studied in the Center of Mass reference frame. In this work we show that this equation can be written in a covariant form in a generic reference frame by using specific momentum variables. These variables are also consistent with the retardless form of the interaction of the model.

Sean Augenstein, Andrew Hard, Lin Ning, Karan Singhal

Federated learning (FL) enables learning from decentralized privacy-sensitive data, with computations on raw data confined to take place at edge clients. This paper introduces mixed FL, which incorporates an additional loss term calculated at the coordinating server (while maintaining FL's private data restrictions). There are numerous benefits. For example,

Dinh-Thuan Do, Anh-Tu Le, Shahid Mumtaz

Since Internet of Things (IoT) is suggested as the fundamental platform to adapt massive connections and secure transmission, we study physical-layer authentication in the point-to-point wireless systems relying on reconfigurable intelligent surfaces (RIS) technique. Due to lack of direct link from IoT devices (both legal and illegal devices) to the access p

Kyle Gilman, Sam Burer, Laura Balzano

We study the maximization of sums of heterogeneous quadratic forms over the Stiefel manifold, a nonconvex problem that arises in several modern signal processing and machine learning applications such as heteroscedastic probabilistic principal component analysis (HPPCA). In this work, we derive a novel semidefinite program (SDP) relaxation of the original pr

Edgar Cortes-Suárez, C. Alenka Negrete, Héctor M. Hernández-Toledo, Héctor Ibarra-Medel

We present a method to identify type-1 active galactic nuclei (AGN) in the central 3 arcsec integrated spectra of galaxies in the MaNGA DR15 sample. It is based on flux ratios estimates in spectral bands flanking the expected H$\alpha$ broad component H$\alpha_{BC}$. The high signal-to-noise ratio obtained (mean S/N = 84) permits the identification of H$\alp

Yik Lun Kei, Yanzhen Chen, Oscar Hernan Madrid Padilla

The Exponential-family Random Graph Model (ERGM) is a powerful model to fit networks with complex structures. However, for dynamic valued networks whose observations are matrices of counts that evolve over time, the development of the ERGM framework is still in its infancy. To facilitate the modeling of dyad value increment and decrement, a Partially Separab

Charlie Mom, Peter van den Besselaar

Bias in grant allocation is a critical issue, as the expectation is that grants are given to the best researchers, and not to applicants that are socially, organizationally, or topic-wise near-by the decision-makers. In this paper, we investigate the effect of organizational proximity, defined as an applicant with the same affiliation as one of the panel mem

Mats G. Löfdahl, Tomas Hillberg

Images collected with ground-based telescopes suffer blurring and distortions from turbulence in Earth's atmosphere. Adaptive optics (AO) can only partially compensate for these effects. Neither multi-frame blind deconvolution (MFBD) nor speckle techniques restore AO-compensated images to the correct power spectrum and contrast. MFBD can only compensate for

F. Kunzweiler, B. Biltzinger, J. Greiner, J. M. Burgess

In the era of time-domain, multi-messenger astronomy, the detection of transient events on the high-energy electromagnetic sky has become more important than ever. Previous attempts to systematically search for onboard-untriggered events in the data of Fermi-GBM have been limited to short-duration signals with variability time scales smaller than ~1 min due

Shiqiang Wang, Mingyue Ji

Federated learning (FL) faces challenges of intermittent client availability and computation/communication efficiency. As a result, only a small subset of clients can participate in FL at a given time. It is important to understand how partial client participation affects convergence, but most existing works have either considered idealized participation pat

Emmanuel Abbe, Samy Bengio, Elisabetta Cornacchia, Jon Kleinberg

This paper considers the Pointer Value Retrieval (PVR) benchmark introduced in [ZRKB21], where a 'reasoning' function acts on a string of digits to produce the label. More generally, the paper considers the learning of logical functions with gradient descent (GD) on neural networks. It is first shown that in order to learn logical functions with gradient des

Nyle Siddiqui, Rushit Dave, Naeem Seliya, Mounika Vanamala

Static authentication methods, like passwords, grow increasingly weak with advancements in technology and attack strategies. Continuous authentication has been proposed as a solution, in which users who have gained access to an account are still monitored in order to continuously verify that the user is not an imposter who had access to the user credentials.

Sayle Sigarreta, Sayli Sigarreta, Hugo Cruz-Suarez

In this paper, we study topological indices in random spiro chains via a martingale approach. In which their explicit analytical expressions of the exact distribution, expected value and variance are obtained. As n goes to infinity, the asymptotic normality of topological indices of a random spiro chain is established through the Martingale Central Limit The

Lu Zhang, Xiaowei Yu, Yanjun Lyu, Zhengwang Wu

Effective representation of brain anatomical architecture is fundamental in understanding brain regularity and variability. Despite numerous efforts, it is still difficult to infer reliable anatomical correspondence at finer scale, given the tremendous individual variability in cortical folding patterns. It is even more challenging to disentangle common and

Zizhou Huang, Davi Colli Tozoni, Arvi Gjoka, Zachary Ferguson

We introduce a general differentiable solver for time-dependent deformation problems with contact and friction. Our approach uses a finite element discretization with a high-order time integrator coupled with the recently proposed incremental potential contact method for handling contact and friction forces to solve ODE- and PDE-constrained optimization prob

Hui-Yu Xing, Zhen-Ni Xu, Zhu-Fang Cui, Craig D. Roberts

Using a symmetry-preserving regularisation of a vector$\times$vector contact interaction (SCI), we complete a systematic treatment of twelve semileptonic transitions with vector meson final states: $D\to \rho$, $D_{(s)}\to K^\ast$, $D_s\to \phi$, $B\to \rho$, $B_s\to K^\ast$, $B_{(s)}\to D_{(s)}^\ast$, $B_c \to B_{(s)}^\ast, J/\psi, D^\ast$; and thereby fina

Peter van den Besselaar, Charlie Mom

Gender bias in grant allocation is a deviation from the principle that scientific merit should guide grant decisions. However, most studies on gender bias in grant allocation focus on gender differences in success rates, without including variables that measure merit. This study has two main contributions. Firstly, it includes several merit variables in the

Eloy Geenjaar, Amrit Kashyap, Noah Lewis, Robyn Miller

Functional magnetic resonance imaging (fMRI) data contain complex spatiotemporal dynamics, thus researchers have developed approaches that reduce the dimensionality of the signal while extracting relevant and interpretable dynamics. Models of fMRI data that can perform whole-brain discovery of dynamical latent factors are understudied. The benefits of approa

Syed Hasib Akhter Faruqui, Adel Alaeddini, Jing Wang, Susan P Fisher-Hoch

Multiple chronic conditions (MCC) are one of the biggest challenges of modern times. The evolution of MCC follows a complex stochastic process that is influenced by a variety of risk factors, ranging from pre-existing conditions to modifiable lifestyle behavioral factors (e.g. diet, exercise habits, tobacco use, alcohol use, etc.) to non-modifiable socio-dem

Mike Freedman

In both quantum computing and black hole physics, it is natural to regard some deformations, infinitesimal unitaries, as \emph{easy} and others as \emph{hard}. This has lead to a renewed examination of right-invariant metrics on $\operatorname{SU}(2^N)$. It has been hypothesized that there is a critical such metric -- in the sense of phase transitions -- and

Wan-Jin Yeo, Yao-Rui Yeo, Samu Taulu

This paper reviews magnetic flux signal calculations through pick-up loops using vector spherical harmonic expansion under the quasi-static approximation, and presents a near-analytical method of evaluating the flux through arbitrary parametrizable pick-up loops for each expansion degree. This is done by simplifying the surface flux integral (2D) into a line

Ximing Lu, Sean Welleck, Jack Hessel, Liwei Jiang

Large-scale language models often learn behaviors that are misaligned with user expectations. Generated text may contain offensive or toxic language, contain significant repetition, or be of a different sentiment than desired by the user. We consider the task of unlearning these misalignments by fine-tuning the language model on signals of what not to do. We

Alireza Aghasi, MohammadJavad Feizollahi, Saeed Ghadimi

We present a robust framework to perform linear regression with missing entries in the features. By considering an elliptical data distribution, and specifically a multivariate normal model, we are able to conditionally formulate a distribution for the missing entries and present a robust framework, which minimizes the worst case error caused by the uncertai

Yuhao Zhang, Aws Albarghouthi, Loris D'Antoni

Machine learning models are vulnerable to data-poisoning attacks, in which an attacker maliciously modifies the training set to change the prediction of a learned model. In a trigger-less attack, the attacker can modify the training set but not the test inputs, while in a backdoor attack the attacker can also modify test inputs. Existing model-agnostic defen

Muhammad Umar B. Niazi, Xiaodong Cheng, Carlos Canudas-de-Wit, Jacquelien M. A. Scherpen

This paper addresses the aggregated monitoring problem for large-scale network systems with a few dedicated sensors. Full state estimation of such systems is often infeasible due to unobservability and/or computational infeasibility. Therefore, through clustering and aggregation, a tractable representation of a network system, called a projected network syst

Jai Kwan Bae, Matthew Andorf, Adam Bartnik, Alice Galdi

Negative Electron Affinity (NEA) activated GaAs photocathodes are the most popular option for generating a high current (> 1 mA) spin-polarized electron beam. Despite its popularity, a short operational lifetime is the main drawback of this material. Recent works have shown that the lifetime can be improved by using a robust Cs-Sb-O NEA layer with minimal ad

Pedro A. Santos-Florez, Howard Yanxon, Byungkyun Kang, Yansun Yao

In this work, we present an efficient framework that combines machine learning potential (MLP) and metadynamics to explore multi-dimensional free energy surfaces for investigating solid-solid phase transition. Based on the spectral descriptors and neural networks regression, we have developed a computationally scalable MLP model to warrant an accurate interp

Barbara Lopez-Doriga, Scott T. M. Dawson, Ricardo Vinuesa

This work applies resolvent analysis to incompressible flow through a rectangular duct, in order to identify dominant linear energy-amplification mechanisms present in such flows. In particular, we formulate the resolvent operator from linearizing the Navier--Stokes equations about a two-dimensional base/mean flow. The laminar base flow only has a nonzero st

Hannah Schieber, Fabian Duerr, Torsten Schoen, Jürgen Beyerer

Robust environment perception for autonomous vehicles is a tremendous challenge, which makes a diverse sensor set with e.g. camera, lidar and radar crucial. In the process of understanding the recorded sensor data, 3D semantic segmentation plays an important role. Therefore, this work presents a pyramid-based deep fusion architecture for lidar and camera to

Emil Brinch Holm, Thomas Tram, Steen Hannestad

Decaying dark matter models provide a physically motivated way of channeling energy between the matter and radiation sectors. In principle, this could affect the predicted value of the Hubble constant in such a way as to accommodate the discrepancies between CMB inferences and local measurements of the same. Here, we revisit the model of warm dark matter dec

Mojmír Mutný, Andreas Krause

Optimal experimental design seeks to determine the most informative allocation of experiments to infer an unknown statistical quantity. In this work, we investigate the optimal design of experiments for {\em estimation of linear functionals in reproducing kernel Hilbert spaces (RKHSs)}. This problem has been extensively studied in the linear regression setti

Raphael Sonabend, Florian Pfisterer, Alan Mishler, Moritz Schauer

Algorithmic fairness is an increasingly important field concerned with detecting and mitigating biases in machine learning models. There has been a wealth of literature for algorithmic fairness in regression and classification however there has been little exploration of the field for survival analysis. Survival analysis is the prediction task in which one a

Tianhao Shao, Jinhui Chen, Yu-Gang Ma, Zhangbu Xu

Light antinucleus yields are calculated in a multiphase transport model (AMPT) coupled with a dynamical coalescence model. The model is tuned to reproduce the transverse momentum and rapidity distributions of antiproton in $pp$ collisions at $\rm \sqrt{s}$ = 7.7 GeV to 7 TeV. By applying a widely used cosmic ray propagation model, the antinucleus fluxes near

Sandhya Devi, Sherman Page

In an earlier study, we showed that Tsallis relative entropy (TRE), which is the generalization of Kullback-Leibler relative entropy (KLRE) to non-extensive systems, can be used as a possible risk measure in constructing risk optimal portfolios whose returns beat market returns. Over a long term (> 10 years), the risk-return profiles from TRE as the risk mea

Nima Dehmamy, Csaba Both, Jianzhi Long, Rose Yu

In mathematical optimization, second-order Newton's methods generally converge faster than first-order methods, but they require the inverse of the Hessian, hence are computationally expensive. However, we discover that on sparse graphs, graph neural networks (GNN) can implement an efficient Quasi-Newton method that can speed up optimization by a factor of 1

Jacob Granley, Lucas Relic, Michael Beyeler

Sensory neuroprostheses are emerging as a promising technology to restore lost sensory function or augment human capabilities. However, sensations elicited by current devices often appear artificial and distorted. Although current models can predict the neural or perceptual response to an electrical stimulus, an optimal stimulation strategy solves the invers

Benjamin Stölzner, Benjamin Joachimi, Andreas Korn, the LSST Dark Energy Science Collaboration

We present an optimisation method for the assignment of photometric galaxies into a chosen set of redshift bins. This is achieved by combining simulated annealing, an optimisation algorithm inspired by solid-state physics, with an unsupervised machine learning method, a self-organising map (SOM) of the observed colours of galaxies. Starting with a sample of

Jimit Majmudar, Christophe Dupuy, Charith Peris, Sami Smaili

Recent large-scale natural language processing (NLP) systems use a pre-trained Large Language Model (LLM) on massive and diverse corpora as a headstart. In practice, the pre-trained model is adapted to a wide array of tasks via fine-tuning on task-specific datasets. LLMs, while effective, have been shown to memorize instances of training data thereby potenti

John W. Montano, Hengxiao Guo, Aaron J. Barth, Vivian U

The nearby dwarf spiral galaxy NGC 4395 contains a broad-lined active galactic nucleus (AGN) of exceptionally low luminosity powered by accretion onto a central black hole of very low mass ($\sim10^4-10^5$ M$_\odot$). In order to constrain the size of the optical continuum emission region through reverberation mapping, we carried out high-cadence photometric

Yunqi Li, Hanxiong Chen, Shuyuan Xu, Yingqiang Ge

As one of the most pervasive applications of machine learning, recommender systems are playing an important role on assisting human decision making. The satisfaction of users and the interests of platforms are closely related to the quality of the generated recommendation results. However, as a highly data-driven system, recommender system could be affected

Avishag Shapira, Alon Zolfi, Luca Demetrio, Battista Biggio

Adversarial attacks against deep learning-based object detectors have been studied extensively in the past few years. Most of the attacks proposed have targeted the model's integrity (i.e., caused the model to make incorrect predictions), while adversarial attacks targeting the model's availability, a critical aspect in safety-critical domains such as autono

Aditya Gopalan, Gugan Thoppe

A primary requirement for any reinforcement learning method is that it should produce policies that improve upon the initial guess. In this work, we show that the widely used Deep Q-Network (DQN) fails to satisfy this minimal criterion -- even when it gets to see all possible states and actions infinitely often (a condition under which tabular Q-learning is

Xiangyu Qi, Tinghao Xie, Jiachen T. Wang, Tong Wu

Adversaries can embed backdoors in deep learning models by introducing backdoor poison samples into training datasets. In this work, we investigate how to detect such poison samples to mitigate the threat of backdoor attacks. First, we uncover a post-hoc workflow underlying most prior work, where defenders passively allow the attack to proceed and then lever

Vadim A. Kaimanovich, Wolfgang Woess

We study branching Markov chains on a countable state space (space of types) $\mathscr{X}$, with the focus on the qualitative aspects of the limit behaviour of the evolving empirical population distributions. No conditions are imposed on the multitype offspring distributions at the points of $\mathscr{X}$ other than to have the same average and to satisfy a

Leon Lufkin, Ashish Puri, Ganlin Song, Xinyi Zhong

Local patterns of excitation and inhibition that can generate neural waves are studied as a computational mechanism underlying the organization of neuronal tunings. Sparse coding algorithms based on networks of excitatory and inhibitory neurons are proposed that exhibit topographic maps as the receptive fields are adapted to input stimuli. Motivated by a lea

Xiangyu Qi, Tinghao Xie, Yiming Li, Saeed Mahloujifar

Recent studies revealed that deep learning is susceptible to backdoor poisoning attacks. An adversary can embed a hidden backdoor into a model to manipulate its predictions by only modifying a few training data, without controlling the training process. Currently, a tangible signature has been widely observed across a diverse set of backdoor poisoning attack

Gilad Gour

We find necessary and sufficient conditions to determine the inter-convertibility of quantum systems under time-translation covariant evolution, and use it to solve several problems in quantum thermodynamics both in the single-shot and asymptotic regimes. It is well known that the resource theory of quantum athermality is not reversible, but in PRL 111, 2504

Mikhail R. Gabdullin

Let $\gamma_0=\frac{\sqrt5-1}{2}=0.618\ldots$ . We prove that, for any $\varepsilon>0$ and any trigonometric polynomial $f$ with frequencies in the set $\{n^2: N \leqslant n\leqslant N+N^{\gamma_0-\varepsilon}\}$, the inequality $$ \|f\|_4 \ll \varepsilon^{-1/4}\|f\|_2 $$ holds, which makes a progress on a conjecture of Cilleruelo and Cordoba. We also presen

Moein Naseri, Tulja Varun Kondra, Suchetana Goswami, Marco Fellous-Asiani

Quantum algorithms allow to outperform their classical counterparts in various tasks, most prominent example being Shor's algorithm for efficient prime factorization on a quantum computer. It is clear that one of the reasons for the speedup is the superposition principle of quantum mechanics, which allows a quantum processor to be in a superposition of diffe

Camosso Simone

In this brief article we try to find an ''interpretation'' for the formalism $\sqrt{dx}$.

Zeyu Bian, Erica EM Moodie, Susan M Shortreed, Sylvie D Lambert

An individualized treatment rule (ITR) is a decision rule that aims to improve individual patients health outcomes by recommending optimal treatments according to patients specific information. In observational studies, collected data may contain many variables that are irrelevant for making treatment decisions. Including all available variables in the stati

Adam B Kashlak, Prachi Loliencar, Giseon Heo

The hidden Markov model (HMM) is a classic modeling tool with a wide swath of applications. Its inception considered observations restricted to a finite alphabet, but it was quickly extended to multivariate continuous distributions. In this article, we further extend the HMM from mixtures of normal distributions in $d$-dimensional Euclidean space to general

Mike A. Merrill, Tim Althoff

Detailed mobile sensing data from phones, watches, and fitness trackers offer an unparalleled opportunity to quantify and act upon previously unmeasurable behavioral changes in order to improve individual health and accelerate responses to emerging diseases. Unlike in natural language processing and computer vision, deep representation learning has yet to br

D. A. Zyuzin, S. V. Zharikov, A. V. Karpova, A. Yu. Kirichenko

The 840 kyr old pulsar PSR J1957+5033, detected so far only in $\gamma$- and X-rays, is a nearby and rather cool neutron star with a temperature of 0.2--0.3 MK, a distance of $\la$1 kpc, and a small colour reddening excess $E(B-V) \approx 0.03$. These properties make it an ideal candidate to detect in the optical to get additional constraints on its paramete

Olivia Dumitrescu, Rick Miranda

We investigate the study of smooth irreducible rational curves in $Y_s^r$, a general blowup of $\mathbb{P}^r$ at $s$ general points, whose normal bundle splits as a direct sum of line bundles all of degree $i$, for $i \in \{-1,0,1\}$: we call these $(i)$-curves. We systematically exploit the theory of Coxeter groups applied to the Chow space of curves in $Y_

Carlos M. Correa, Dante J. Paz

Cosmic voids constitute promising cosmological laboratories. However, a full description of all the redshift-space effects that affect observational measurements is mandatory in order to obtain unbiased cosmological constraints. We make a description in a nutshell of these effects and lay the theoretical foundations for designing reliable cosmological tests

Junru Shao, Xiyou Zhou, Siyuan Feng, Bohan Hou

Automatic optimization for tensor programs becomes increasingly important as we deploy deep learning in various environments, and efficient optimization relies on a rich search space and effective search. Most existing efforts adopt a search space which lacks the ability to efficiently enable domain experts to grow the search space. This paper introduces Met

Michael Whitehouse, Nick Whiteley, Lorenzo Rimella

Addressing the challenge of scaling-up epidemiological inference to complex and heterogeneous models, we introduce Poisson Approximate Likelihood (PAL) methods. In contrast to the popular ODE approach to compartmental modelling, in which a large population limit is used to motivate a deterministic model, PALs are derived from approximate filtering equations