arXiv subjects

Xu Zhang

Publications and source records attributed to Xu Zhang.

At least 343 records · Page 19Linked to original sources

arXiv2023

Sparse Personalized Federated Learning

Federated Learning (FL) is a collaborative machine learning technique to train a global model without obtaining clients' private data. The main challenges in FL are statistical diversity among clients, limited computing capability among clients' equipments, and the excessive communication overhead between the server and clients. To address these challenges, we propose a novel sparse personalized federated learning scheme via maximizing correlation (FedMac). By incorporating an approximated L1-norm and the correlation between client models and global model into standard FL loss function, the performance on statistical diversity data is improved and the communicational and computational loads required in the network are reduced compared with non-sparse FL. Convergence analysis shows that the sparse constraints in FedMac do not affect the convergence rate of the global model, and theoretical results show that FedMac can achieve good sparse personalization, which is better than the personalized methods based on L2-norm. Experimentally, we demonstrate the benefits of this sparse personalization architecture compared with the state-of-the-art personalization methods (e.g. FedMac respectively achieves 98.95%, 99.37%, 90.90%, 89.06% and 73.52% accuracy on the MNIST, FMNIST, CIFAR-100, Synthetic and CINIC-10 datasets under non-i.i.d. variants).

cs.LG↗

arXiv2023

Avoiding spurious correlations via logit correction

Empirical studies suggest that machine learning models trained with empirical risk minimization (ERM) often rely on attributes that may be spuriously correlated with the class labels. Such models typically lead to poor performance during inference for data lacking such correlations. In this work, we explicitly consider a situation where potential spurious correlations are present in the majority of training data. In contrast with existing approaches, which use the ERM model outputs to detect the samples without spurious correlations and either heuristically upweight or upsample those samples, we propose the logit correction (LC) loss, a simple yet effective improvement on the softmax cross-entropy loss, to correct the sample logit. We demonstrate that minimizing the LC loss is equivalent to maximizing the group-balanced accuracy, so the proposed LC could mitigate the negative impacts of spurious correlations. Our extensive experimental results further reveal that the proposed LC loss outperforms state-of-the-art solutions on multiple popular benchmarks by a large margin, an average 5.5\% absolute improvement, without access to spurious attribute labels. LC is also competitive with oracle methods that make use of the attribute labels. Code is available at https://github.com/shengliu66/LC.

cs.LG↗

arXiv2023

Robust one-shot estimation over shared networks in the presence of denial-of-service attacks

Multi-agent systems often communicate over low-power shared wireless networks in unlicensed spectrum, prone to denial-of-service attacks. We consider the following scenario: multiple pairs of agents communicating strategically over shared communication networks in the presence of a jammer who may launch a denial-of-service. We cast this problem as a game between a coordinator who optimizes the transmission and estimation policies jointly and a jammer who optimizes its probability of performing an attack. We consider two cases: point-to-point channels and large-scale networks with a countably infinite number of sensor-receiver pairs. When the jammer proactively attacks the channel, the game is nonconvex from the coordinator's perspective. However, despite the lack of convexity, we construct a saddle point equilibrium solution for any multi-variate Gaussian distribution for the observations. When the jammer is reactive, we obtain an algorithm based on sequential convex optimization, which converges swiftly to first-order Nash-equilibria. Interestingly, blocking the channel is often optimal when the jammer is reactive, even when it is idle, to create ambiguity at the receiver.

eess.SY↗

arXiv2023

Incident-aware Duplicate Ticket Aggregation for Cloud Systems

In cloud systems, incidents are potential threats to customer satisfaction and business revenue. When customers are affected by incidents, they often request customer support service (CSS) from the cloud provider by submitting a support ticket. Many tickets could be duplicate as they are reported in a distributed and uncoordinated manner. Thus, aggregating such duplicate tickets is essential for efficient ticket management. Previous studies mainly rely on tickets' textual similarity to detect duplication; however, duplicate tickets in a cloud system could carry semantically different descriptions due to the complex service dependency of the cloud system. To tackle this problem, we propose iPACK, an incident-aware method for aggregating duplicate tickets by fusing the failure information between the customer side (i.e., tickets) and the cloud side (i.e., incidents). We extensively evaluate iPACK on three datasets collected from the production environment of a large-scale cloud platform, Azure. The experimental results show that iPACK can precisely and comprehensively aggregate duplicate tickets, achieving an F1 score of 0.871 ~ 0.935 and outperforming state-of-the-art methods by 12.4% ~ 31.2%.

cs.SE↗

arXiv2023

Optical bulk-boundary dichotomy in a quantum spin Hall insulator

The bulk-boundary correspondence is a key concept in topological quantum materials. For instance, a quantum spin Hall insulator features a bulk insulating gap with gapless helical boundary states protected by the underlying Z2 topology. However, the bulk-boundary dichotomy and distinction are rarely explored in optical experiments, which can provide unique information about topological charge carriers beyond transport and electronic spectroscopy techniques. Here, we utilize mid-infrared absorption micro-spectroscopy and pump-probe micro-spectroscopy to elucidate the bulk-boundary optical responses of Bi4Br4, a recently discovered room-temperature quantum spin Hall insulator. Benefiting from the low energy of infrared photons and the high spatial resolution, we unambiguously resolve a strong absorption from the boundary states while the bulk absorption is suppressed by its insulating gap. Moreover, the boundary absorption exhibits a strong polarization anisotropy, consistent with the one-dimensional nature of the topological boundary states. Our infrared pump-probe microscopy further measures a substantially increased carrier lifetime for the boundary states, which reaches one nanosecond scale. The nanosecond lifetime is about one to two orders longer than that of most topological materials and can be attributed to the linear dispersion nature of the helical boundary states. Our findings demonstrate the optical bulk-boundary dichotomy in a topological material and provide a proof-of-principal methodology for studying topological optoelectronics.

cond-mat.mes-hall↗

arXiv2023

Evidence for gapless quantum spin liquid in a honeycomb lattice

One main theme in current condensed matter physics is the search of quantum spin liquid (QSL), an exotic magnetic state with strongly-fluctuating and highly-entangled spins down to zero temperature without static order. However, there is no consensus on the existence of a QSL ground state in any real material so far. The disorders and competing exchange interactions may prevent the formation of an ideal QSL state on frustrated spin lattices. Here we report systematic heat transport measurements on a honeycomb-lattice compound BaCo2(AsO4)2, which manifests magnetic order in zero field. In a narrow field range after the magnetic order is nearly suppressed by an in-plane field, in both perpendicular and parallel to the zigzag direction, a finite residual linear term of thermal conductivity is clearly observed, which is attributed to the mobile fractionalized spinon excitations. This provides smoking-gun evidence for a gapless QSL state in BaCo2(AsO4)2. We discuss the underlying physics to form this exotic gapless QSL state in Co2+ honeycomb lattice.

cond-mat.str-el↗

arXiv2023

Online Decomposition of Surface Electromyogram into Individual Motor Unit Activities Using Progressive FastICA Peel-off

Surface electromyogram (SEMG) decomposition provides a promising tool for decoding and understanding neural drive information non-invasively. In contrast to previous SEMG decomposition methods mainly developed in offline conditions, there are few studies on online SEMG decomposition. A novel method for online decomposition of SEMG data is presented using the progressive FastICA peel-off (PFP) algorithm. The online method consists of an offline prework stage and an online decomposition stage. More specifically, a series of separation vectors are first initialized by the originally offline version of the PFP algorithm from SEMG data recorded in advance. Then they are applied to online SEMG data to extract motor unit spike trains precisely. The performance of the proposed online SEMG decomposition method was evaluated by both simulation and experimental approaches. It achieved an online decomposition accuracy of 98.53% when processing simulated SEMG data. For decomposing experimental SEMG data, the proposed online method was able to extract an average of 12.00 +- 3.46 MUs per trial, with a matching rate of 90.38% compared with results from the expert-guided offline decomposition. Our study provides a valuable way of online decomposition of SEMG data with advanced applications in movement control and health.

eess.SP↗

arXiv2023

In-situ electrical and thermal transport properties of FeySe1-xTex films with ionic liquid gating

We combine in-situ electrical transport and Seebeck coefficient measurements with the ionic liquid gating technique to investigate superconductivity and the normal state of FeySe1-xTex (FST) films. We find that the pristine FST films feature a non-Fermi liquid temperature dependence of the Seebeck coefficient, i.e., S/T ~ AS lnT, and AS is strongly correlated with the superconducting transition temperature (Tc). Ionic liquid gating significantly raises Tc of FST films, for which the Seebeck coefficient displays a novel scaling behavior and retains the logarithmic temperature dependence. Moreover, a quantitative relationship between the slope of T-linear resistivity (A\r{ho}) and Tc for gated films is observed, i.e., (A\r{ho})1/2 ~ Tc, consistent with previous reports on cuprates and FeSe. The scaling behaviors of AS and A\r{ho} point to a spin-fluctuation-associated transport mechanism in gated FeySe1-xTex superconductors.

cond-mat.supr-con↗

arXiv2023

Surveillance Face Anti-spoofing

Face Anti-spoofing (FAS) is essential to secure face recognition systems from various physical attacks. However, recent research generally focuses on short-distance applications (i.e., phone unlocking) while lacking consideration of long-distance scenes (i.e., surveillance security checks). In order to promote relevant research and fill this gap in the community, we collect a large-scale Surveillance High-Fidelity Mask (SuHiFiMask) dataset captured under 40 surveillance scenes, which has 101 subjects from different age groups with 232 3D attacks (high-fidelity masks), 200 2D attacks (posters, portraits, and screens), and 2 adversarial attacks. In this scene, low image resolution and noise interference are new challenges faced in surveillance FAS. Together with the SuHiFiMask dataset, we propose a Contrastive Quality-Invariance Learning (CQIL) network to alleviate the performance degradation caused by image quality from three aspects: (1) An Image Quality Variable module (IQV) is introduced to recover image information associated with discrimination by combining the super-resolution network. (2) Using generated sample pairs to simulate quality variance distributions to help contrastive learning strategies obtain robust feature representation under quality variation. (3) A Separate Quality Network (SQN) is designed to learn discriminative features independent of image quality. Finally, a large number of experiments verify the quality of the SuHiFiMask dataset and the superiority of the proposed CQIL.

cs.CV↗

arXiv2022

Top-k data selection via distributed sample quantile inference

We consider the problem of determining the top-$k$ largest measurements from a dataset distributed among a network of $n$ agents with noisy communication links. We show that this scenario can be cast as a distributed convex optimization problem called sample quantile inference, which we solve using a two-time-scale stochastic approximation algorithm. Herein, we prove the algorithm's convergence in the almost sure sense to an optimal solution. Moreover, our algorithm handles noise and empirically converges to the correct answer within a small number of iterations.

cs.DC↗

arXiv2022

Superconductivity and bosonic fluid emerging from Moiré flat bands

Although evidence of inter-valley attraction-mediated by phonon or topological fluctuations is accumulating, the origin of superconductivity in the flat-band quantum moiré materials remains an open question. Here, instead of attempting to pinpoint the origin of the superconductivity, we aim at identifying universal properties of moiré flat bands that shall emerge in the presence of inter-valley attractions. We show that by matching the interaction strength of inter-valley attraction with intra-valley repulsion, the flat-band limit becomes exactly solvable. Away from the flat-band limit, the system can be simulated via quantum Monte Carlo (QMC) methods without sign problem for any fillings. Combining analytic solutions with large-scale numerical simulations, we show that upon increasing temperature, the superconducting phase melts into a bosonic fluid of Cooper pairs with large/diverging compressibility. In contrast to flat-band attractive Hubbard models, where similar effects arise only for on-site interactions, our study indicates this physics is a universal property of moiré flat bands, regardless of microscopic details such as the range of interactions and/or spin-oribt couplings. At higher temperature, the boson fluid phase gives its way to a pseudo gap phase, where some Cooper pairs are torn apart by thermal fluctuations, resulting in fermion density of states inside the gap. Unlike the superconducting transition temperature, which is very sensitive to doping and twisting angles, the gap and the temperature scale of the boson fluid phase and the pseudo gap phase are found to be nearly independent of doping level and/or flat-band bandwidth. The relevance of these phases with experimental discoveries in the flat band quantum moiré materials is discussed.

cond-mat.supr-con↗

arXiv2022

Enabling Personalized Video Quality Optimization with VidHoc

The emerging video applications greatly increase the demand in network bandwidth that is not easy to scale. To provide higher quality of experience (QoE) under limited bandwidth, a recent trend is to leverage the heterogeneity of quality preferences across individual users. Although these efforts have suggested the great potential benefits, service providers still have not deployed them to realize the promised QoE improvement. The missing piece is an automation of online per-user QoE modeling and optimization scheme for new users. Previous efforts either optimize QoE by known per-user QoE models or learn a user's QoE model by offline approaches, such as analysis of video viewing history and in-lab user study. Relying on such offline modeling is problematic, because QoE optimization will start late for collecting enough data to train an unbiased QoE model. In this paper, we propose VidHoc, the first automatic system that jointly personalizes QoE model and optimizes QoE in an online manner for each new user. VidHoc can build per-user QoE models within a small number of video sessions as well as maintain good QoE. We evaluate VidHoc in a pilot deployment to fifteen users for four months with the care of statistical validity. Compared with other baselines, the results show that VidHoc can save 17.3% bandwidth while maintaining the same QoE or improve QoE by 13.9% with the same bandwidth.

cs.NI↗

arXiv2022

Umehara algebra and complex submanifolds of indefinite complex space forms

The Umehara algebra is studied with motivation on the problem of the non-existence of common complex submanifolds. In this paper, we prove some new results in Umehara algebra and obtain some applications. In particular, if a complex manifolds admits a holomorphic polynomial isometric immersion to one indefinite complex space form, then it cannot admits a holomorphic isometric immersion to another indefinite complex space form of different type. Other consequences include the non-existence of the common complex submanifolds for indefinite complex projective space or hyperbolic space and a complex manifold with a distinguished metric, such as homogeneous domains, the Hartogs triangle, the minimal ball, the symmetrized polydisc, etc, equipped with their intrinsic Bergman metrics, which generalizes more or less all existing results.

math.CV↗

arXiv2022

Quantum Monte Carlo sign bounds, topological Mott insulator and thermodynamic transitions in twisted bilayer graphene model

We show that for magic-angle twisted bilayer graphene (TBG) away from charge neutrality, although quantum Monte Carlo (QMC) simulations suffer from the sign problem, the computational complexity is at most polynomial at certain integer fillings. For even integer fillings, this polynomial complexity survives even if an extra inter-valley attractive interaction is introduced, on top of Coulomb repulsions. This observation allows us to simulate magic-angle twisted bilayer graphene and to obtain accurate phase diagram and dynamical properties. At the chiral limit and filling $ν=1$, the simulations reveal a thermodynamic transition separating metallic state and a $C=1$ correlated Chern insulator -- topological Mott insulator (TMI) -- and the pseudogap spectrum slightly above the transition temperature. The ground state excitation spectra of the TMI exhibit a spin-valley U(4) Goldstone mode and a time reversal restoring excitonic gap smaller than the single particle gap. These results are qualitatively consistent with the recent experimental findings at zero-field and $ν=1$ filling in $h$-BN nonaligned TBG.

cond-mat.str-el↗

arXiv2022

Low-rank Latent Matrix-factor Prediction Modeling for Generalized High-dimensional Matrix-variate Regression

Motivated by diagnosing the COVID-19 disease using 2D image biomarkers from computed tomography (CT) scans, we propose a novel latent matrix-factor regression model to predict responses that may come from an exponential distribution family, where covariates include high-dimensional matrix-variate biomarkers. A latent generalized matrix regression (LaGMaR) is formulated, where the latent predictor is a low-dimensional matrix factor score extracted from the low-rank signal of the matrix variate through a cutting-edge matrix factor model. Unlike the general spirit of penalizing vectorization plus the necessity of tuning parameters in the literature, instead, our prediction modeling in LaGMaR conducts dimension reduction that respects the geometry characteristic of intrinsic two-dimensional structure of the matrix covariate and thus avoids iteration. This greatly relieves the computation burden, and meanwhile maintains structural information so that the latent matrix factor feature can perfectly replace the intractable matrix-variate owing to high-dimensionality. The estimation procedure of LaGMaR is subtly derived by transforming the bilinear form matrix factor model onto a high-dimensional vector factor model, so that the method of principle components can be applied. We establish bilinear-form consistency of the estimated matrix coefficient of the latent predictor and consistency of prediction. The proposed approach can be implemented conveniently. Through simulation experiments, prediction capability of LaGMaR is shown to outperform existing penalized methods under diverse scenarios of generalized matrix regressions. Through the application to a real COVID-19 dataset, the proposed approach is shown to predict efficiently the COVID-19.

stat.AP↗

arXiv2022

Empirical likelihood-based portmanteau tests for autoregressive moving average models with possible infinite variance innovation

It is an important task in the literature to check whether a fitted autoregressive moving average (ARMA) model is adequate, while the currently used tests may suffer from the size distortion problem when the underlying autoregressive models have low persistence. To fill this gap, this paper proposes two empirical likelihood-based portmanteau tests. The first one is naive but can serve as a benchmark, and the second is for the case with infinite variance innovations. The asymptotic distributions under the null hypothesis are derived under mild moment conditions, and their usefulness is demonstrated by simulation experiments and two real data examples.

stat.ME↗

arXiv2022

An Empirical Study of Automatic Post-Editing

Automatic post-editing (APE) aims to reduce manual post-editing efforts by automatically correcting errors in machine-translated output. Due to the limited amount of human-annotated training data, data scarcity is one of the main challenges faced by all APE systems. To alleviate the lack of genuine training data, most of the current APE systems employ data augmentation methods to generate large-scale artificial corpora. In view of the importance of data augmentation in APE, we separately study the impact of the construction method of artificial corpora and artificial data domain on the performance of APE models. Moreover, the difficulty of APE varies between different machine translation (MT) systems. We study the outputs of the state-of-art APE model on a difficult APE dataset to analyze the problems in existing APE systems. Primarily, we find that 1) Artificial corpora with high-quality source text and machine-translated text more effectively improve the performance of APE models; 2) In-domain artificial training data can better improve the performance of APE models, while irrelevant out-of-domain data actually interfere with the model; 3) Existing APE model struggles with cases containing long source text or high-quality machine-translated text; 4) The state-of-art APE model works well on grammatical and semantic addition problems, but the output is prone to entity and semantic omission errors.

cs.CL↗

arXiv2022

Rate-Distortion Theory for Strategic Semantic Communication

This paper analyzes the fundamental limit of the strategic semantic communication problem in which a transmitter obtains a limited number of indirect observation of an intrinsic semantic information source and can then influence the receiver's decoding by sending a limited number of messages to an imperfect channel. The transmitter and the receiver can have different distortion measures and can make rational decision about their encoding and decoding strategies, respectively. The decoder can also have some side information (e.g., background knowledge and/or information obtained from previous communications) about the semantic source to assist its interpretation of the semantic information. We focus particularly on the case that the transmitter can commit to an encoding strategy and study the impact of the strategic decision making on the rate distortion of semantic communication. Three equilibrium solutions including the strong Stackelberg equilibrium, weak Stackelberg equilibrium, as well as Nash equilibrium have been studied and compared. The optimal encoding and decoding strategy profiles under various equilibrium solutions have been derived. We prove that committing to an encoding strategy cannot always bring benefit to the encoder. We therefore propose a feasible condition under which committing to an encoding strategy can always reduce the distortion performance of semantic communication.

cs.IT↗