arXiv subjects

Kai Zhou

Publications and source records attributed to Kai Zhou.

At least 163 records · Page 9Linked to original sources

arXiv2022

BinarizedAttack: Structural Poisoning Attacks to Graph-based Anomaly Detection

Graph-based Anomaly Detection (GAD) is becoming prevalent due to the powerful representation abilities of graphs as well as recent advances in graph mining techniques. These GAD tools, however, expose a new attacking surface, ironically due to their unique advantage of being able to exploit the relations among data. That is, attackers now can manipulate those relations (i.e., the structure of the graph) to allow some target nodes to evade detection. In this paper, we exploit this vulnerability by designing a new type of targeted structural poisoning attacks to a representative regression-based GAD system termed OddBall. Specially, we formulate the attack against OddBall as a bi-level optimization problem, where the key technical challenge is to efficiently solve the problem in a discrete domain. We propose a novel attack method termed BinarizedAttack based on gradient descent. Comparing to prior arts, BinarizedAttack can better use the gradient information, making it particularly suitable for solving combinatorial optimization problems. Furthermore, we investigate the attack transferability of BinarizedAttack by employing it to attack other representation-learning-based GAD systems. Our comprehensive experiments demonstrate that BinarizedAttack is very effective in enabling target nodes to evade graph-based anomaly detection tools with limited attackers' budget, and in the black-box transfer attack setting, BinarizedAttack is also tested effective and in particular, can significantly change the node embeddings learned by the GAD systems. Our research thus opens the door to studying a new type of attack against security analytic tools that rely on graph data.

cs.LG↗

arXiv2021

A Multi-user Oriented Live Free-viewpoint Video Streaming System Based On View Interpolation

As an important application form of immersive multimedia services, free-viewpoint video(FVV) enables users with great immersive experience by strong interaction. However, the computational complexity of virtual view synthesis algorithms poses a significant challenge to the real-time performance of an FVV system. Furthermore, the individuality of user interaction makes it difficult to serve multiple users simultaneously for a system with conventional architecture. In this paper, we novelly introduce a CNN-based view interpolation algorithm to synthesis dense virtual views in real time. Based on this, we also build an end-to-end live free-viewpoint system with a multi-user oriented streaming strategy. Our system can utilize a single edge server to serve multiple users at the same time without having to bring a large view synthesis load on the client side. We analyze the whole system and show that our approaches give the user a pleasant immersive experience, in terms of both visual quality and latency.

cs.MM↗

arXiv2021

Automatic differentiation approach for reconstructing spectral functions with neural networks

Reconstructing spectral functions from Euclidean Green's functions is an important inverse problem in physics. The prior knowledge for specific physical systems routinely offers essential regularization schemes for solving the ill-posed problem approximately. Aiming at this point, we propose an automatic differentiation framework as a generic tool for the reconstruction from observable data. We represent the spectra by neural networks and set chi-square as loss function to optimize the parameters with backward automatic differentiation unsupervisedly. In the training process, there is no explicit physical prior embedding into neural networks except the positive-definite form. The reconstruction accuracy is assessed through Kullback-Leibler(KL) divergence and mean square error(MSE) at multiple noise levels. It should be noted that the automatic differential framework and the freedom of introducing regularization are inherent advantages of the present approach and may lead to improvements of solving inverse problem in the future.

physics.comp-ph↗

arXiv2021

The Compressible Euler and Acoustic Limits from quantum Boltzmann Equation with Fermi-Dirac Statistics

This paper justifies the compressible Euler and acoustic limits from quantum Boltzmann equation with Fermi-Dirac statistics (briefly, BFD) rigorously. This limit was formally derived in Zakrevskiy's thesis \cite{Zakrevskiy} by moment method. We employ the Hilbert approach. The forms of the classical compressible Euler system and the one derived from BFD are different. Our proof is based on the analysis of the nonlinear implicit transformation of these two forms, and a few novel nonlinear estimates.

math.AP↗

arXiv2021

An equation-of-state-meter for CBM using PointNet

A novel method for identifying the nature of QCD transitions in heavy-ion collision experiments is introduced. PointNet based Deep Learning (DL) models are developed to classify the equation of state (EoS) that drives the hydrodynamic evolution of the system created in Au-Au collisions at 10 AGeV. The DL models were trained and evaluated in different hypothetical experimental situations. A decreased performance is observed when more realistic experimental effects (acceptance cuts and decreased resolutions) are taken into account. It is shown that the performance can be improved by combining multiple events to make predictions. The PointNet based models trained on the reconstructed tracks of charged particles from the CBM detector simulation discriminate a crossover transition from a first order phase transition with an accuracy of up to 99.8%. The models were subjected to several tests to evaluate the dependence of its performance on the centrality of the collisions and physical parameters of fluid dynamic simulations. The models are shown to work in a broad range of centralities (b=0-7 fm). However, the performance is found to improve for central collisions (b=0-3 fm). There is a drop in the performance when the model parameters lead to reduced duration of the fluid dynamic evolution or when less fraction of the medium undergoes the transition. These effects are due to the limitations of the underlying physics and the DL models are shown to be superior in its discrimination performance in comparison to conventional mean observables.

hep-ph↗

arXiv2021

Probabilistic Bearing Fault Diagnosis Using Gaussian Process with Tailored Feature Extraction

Rolling bearings are subject to various faults due to its long-time operation under harsh environment, which will lead to unexpected breakdown of machinery system and cause severe accidents. Deep learning methods recently have gained growing interests and extensively applied in the data-driven bearing fault diagnosis. However, current deep learning methods perform the bearing fault diagnosis in the form of deterministic classification, which overlook the uncertainties that inevitably exist in actual practice. To tackle this issue, in this research we develop a probabilistic fault diagnosis framework that can account for the uncertainty effect in prediction, which bears practical significance. This framework fully leverages the probabilistic feature of Gaussian process classifier (GPC). To facilitate the establishment of high-fidelity GPC, the tailored feature extraction with dimensionality reduction method can be optimally determined through the cross validation-based grid search upon a prespecified method pool consisting of various kernel principal component analysis (KPCA) methods and stacked autoencoder. This strategy can ensure the complex nonlinear relations between the features and faults to be adequately characterized. Furthermore, the sensor fusion concept is adopted to enhance the diagnosis performance. As compared with the traditional deep learning methods, this proposed framework usually requires less labeled data and less effort for parameter tuning. Systematic case studies using the publicly accessible experimental rolling bearing dataset are carried out to validate this new framework. Various influencing factors on fault diagnosis performance also are thoroughly investigated.

cs.LG↗

arXiv2021

EPick: Multi-Class Attention-based U-shaped Neural Network for Earthquake Detection and Seismic Phase Picking

Earthquake detection and seismic phase picking not only play a crucial role in travel time estimation of body waves(P and S waves) but also in the localisation of the epicenter of the corresponding event. Generally, manual phase picking is a trustworthy and the optimum method to determine the phase arrival time, however, its capacity is restricted by available resources and time. Moreover, noisy seismic data renders an additional critical challenge for fast and accurate phase picking. In this study, a deep learning based model, EPick, is proposed which benefits both from U shaped neural network (also called UNet)and attention mechanism, as a strong alternative for seismic event detection and phase picking. On one hand, the utilization of UNet structure enables addressing different levels of deep features. On the other hand, attention mechanism promotes the decoder in the UNet structure to focus on the efficient exploitation of the low-resolution features learned from the encoder part to achieve precise phase picking. Extensive experimental results demonstrate that EPick achieves better performance over the benchmark method, and show the models robustness when tested on a different seismic dataset.

physics.geo-ph↗

arXiv2021

Learning Langevin dynamics with QCD phase transition

In this proceeding, the deep Convolutional Neural Networks (CNNs) are deployed to recognize the order of QCD phase transition and predict the dynamical parameters in Langevin processes. To overcome the intrinsic randomness existed in a stochastic process, we treat the final spectra as image-type inputs which preserve sufficient spatiotemporal correlations. As a practical example, we demonstrate this paradigm for the scalar condensation in QCD matter near the critical point, in which the order parameter of chiral phase transition can be characterized in a $1+1$-dimensional Langevin equation for $σ$ field. The well-trained CNNs accurately classify the first-order phase transition and crossover from $σ$ field configurations with fluctuations, in which the noise does not impair the performance of the recognition. In reconstructing the dynamics, we demonstrate it is robust to extract the damping coefficients $η$ from the intricate field configurations.

nucl-th↗

arXiv2021

Finding signatures of the nuclear symmetry energy in heavy-ion collisions with deep learning

A deep convolutional neural network (CNN) is developed to study symmetry energy $E_{\rm sym}(ρ)$ effects by learning the mapping between the symmetry energy and the two-dimensional (transverse momentum and rapidity) distributions of protons and neutrons in heavy-ion collisions. Supervised training is performed with labelled data-set from the ultrarelativistic quantum molecular dynamics (UrQMD) model simulation. It is found that, by using proton spectra on event-by-event basis as input, the accuracy for classifying the soft and stiff $E_{\rm sym}(ρ)$ is about 60% due to large event-by-event fluctuations, while by setting event-summed proton spectra as input, the classification accuracy increases to 98%. The accuracy for 5-label (5 different $E_{\rm sym}(ρ)$) classification task are about 58% and 72% by using proton and neutron spectra, respectively. For the regression task, the mean absolute error (MAE) which measures the average magnitude of the absolute differences between the predicted and actual $L$ (the slope parameter of $E_{\rm sym}(ρ)$) are about 20.4 and 14.8 MeV by using proton and neutron spectra, respectively. Fingerprints of the density-dependent nuclear symmetry energy on the transverse momentum and rapidity distributions of protons and neutrons can be identified by convolutional neural network algorithm.

nucl-th↗

arXiv2021

Detecting Chiral Magnetic Effect via Deep Learning

The search of chiral magnetic effect (CME) in heavy-ion collisions has attracted long-term attentions. Multiple observables have been proposed but all suffer from obstacles due to large background contaminations. In this Letter, we construct an observable-independent CME-meter based on a deep convolutional neural network. After trained over data set generated by a multiphase transport model, the CME-meter shows high accuracy in recognizing the CME-featured charge separation from the final-state pion spectra. It also exhibits remarkable robustness to diverse conditions including different collision energies, centralities, and elliptic flow backgrounds. In a transfer learning manner, the CME-meter is validated in isobaric collision systems, showing good transferability among different colliding systems. Based on variational approaches, we utilize the DeepDream method to derive the most responsive CME-spectra that demonstrates the physical contents the machine learns.

hep-ph↗

arXiv2021

Machine learning based approach to fluid dynamics

We study the applicability of a Deep Neural Network (DNN) approach to simulate one-dimensional non-relativistic fluid dynamics. Numerical fluid dynamical calculations are used to generate training data-sets corresponding to a broad range of profiles to perform supervised learning with DNN. The performance of the DNN approach is analyzed, with a focus on its interpolation and extrapolation capabilities. Issues such as inference speed, the networks capacities to interpolate and extrapolate solutions with limited training samples from both initial geometries and evolution duration aspects are studied in detail. The optimal DNN performance is achieved when its objective is set to learn the mapping between hydro profiles after a fixed value time step, which can then be applied successively to reach moments in time much beyond the duration contained in the training. The DNN has an advantage over the conventional numerical methods by not being restricted by the Courant criterion, and it shows a speedup over the conventional numerical methods by at least two orders of magnitude.

physics.comp-ph↗

arXiv2021

Deep learning stochastic processes with QCD phase transition

It is non-trivial to recognize phase transitions and track dynamics inside a stochastic process because of its intrinsic stochasticity. In this paper, we employ the deep learning method to classify the phase orders and predict the damping coefficient of fluctuating systems under Langevin's description. As a concrete set-up, we demonstrate this paradigm for the scalar condensation in QCD matter near the critical point, in which the order parameter of chiral phase transition can be characterized in a $1+1$-dimensional Langevin equation for $σ$ field. In a supervised learning manner, the Convolutional Neural Networks(CNNs) accurately classify the first-order phase transition and crossover based on $σ$ field configurations with fluctuations. Noise in the stochastic process does not significantly hinder the performance of the well-trained neural network for phase order recognition. For mixed dynamics with diverse dynamical parameters, we further devise and train the machine to predict the damping coefficients $η$ in a broad range. The results show that it is robust to extract the dynamics from the bumpy field configurations.

nucl-th↗

arXiv2021

The Incompressible Navier-Stokes-Fourier Limit from Boltzmann-Fermi-Dirac Equation

We study Boltzmann-Fermi-Dirac equation when quantum effects are taken into account in dilute gas dynamics. By employing new estimates on trilinear terms of collision kernels, we prove the global existence of the classical solution to Boltzmann-Fermi-Dirac equation near equilibrium. Furthermore, the limit from Boltzmann-Fermi-Dirac equation to incompressible Navier-Stokes-Fourier equations is justified rigorously. The corresponding formal analysis was given in the thesis of Zakrevskiy \cite{Zakrevskiy}

math.AP↗

arXiv2021

Strategic Evasion of Centrality Measures

Among the most fundamental tools for social network analysis are centrality measures, which quantify the importance of every node in the network. This centrality analysis typically disregards the possibility that the network may have been deliberately manipulated to mislead the analysis. To solve this problem, a recent study attempted to understand how a member of a social network could rewire the connections therein to avoid being identified as a leader of that network. However, the study was based on the assumption that the network analyzer - the seeker - is oblivious to any evasion attempts by the evader. In this paper, we relax this assumption by modelling the seeker and evader as strategic players in a Bayesian Stackelberg game. In this context, we study the complexity of various optimization problems, and analyze the equilibria of the game under different assumptions, thereby drawing the first conclusions in the literature regarding which centralities the seeker should use to maximize the chances of detecting a strategic evader.

cs.SI↗

arXiv2020

Machine learning spatio-temporal epidemiological model to evaluate Germany-county-level COVID-19 risk

As the COVID-19 pandemic continues to ravage the world, it is of critical significance to provide a timely risk prediction of the COVID-19 in multi-level. To implement it and evaluate the public health policies, we develop a framework with machine learning assisted to extract epidemic dynamics from the infection data, in which contains a county-level spatiotemporal epidemiological model that combines a spatial Cellular Automaton (CA) with a temporal Susceptible-Undiagnosed-Infected-Removed (SUIR) model. Compared with the existing time risk prediction models, the proposed CA-SUIR model shows the multi-level risk of the county to the government and coronavirus transmission patterns under different policies. This new toolbox is first utilized to the projection of the multi-level COVID-19 prevalence over 412 Landkreis (counties) in Germany, including t-day-ahead risk forecast and the risk assessment to the travel restriction policy. As a practical illustration, we predict the situation at Christmas where the worst fatalities are 34.5 thousand, effective policies could contain it to below 21 thousand. Such intervenable evaluation system could help decide on economic restarting and public health policies making in pandemic.

physics.soc-ph↗

arXiv2020

Blocking Adversarial Influence in Social Networks

While social networks are widely used as a media for information diffusion, attackers can also strategically employ analytical tools, such as influence maximization, to maximize the spread of adversarial content through the networks. We investigate the problem of limiting the diffusion of negative information by blocking nodes and edges in the network. We formulate the interaction between the defender and the attacker as a Stackelberg game where the defender first chooses a set of nodes to block and then the attacker selects a set of seeds to spread negative information from. This yields an extremely complex bi-level optimization problem, particularly since even the standard influence measures are difficult to compute. Our approach is to approximate the attacker's problem as the maximum node domination problem. To solve this problem, we first develop a method based on integer programming combined with constraint generation. Next, to improve scalability, we develop an approximate solution method that represents the attacker's problem as an integer program, and then combines relaxation with duality to yield an upper bound on the defender's objective that can be computed using mixed integer linear programming. Finally, we propose an even more scalable heuristic method that prunes nodes from the consideration set based on their degree. Extensive experiments demonstrate the efficacy of our approaches.

cs.SI↗

arXiv2020

A fast centrality-meter for heavy-ion collisions at the CBM experiment

A new method of event characterization based on Deep Learning is presented. The PointNet models can be used for fast, online event-by-event impact parameter determination at the CBM experiment. For this study, UrQMD and the CBM detector simulation are used to generate Au+Au collision events at 10 AGeV which are then used to train and evaluate PointNet based architectures. The models can be trained on features like the hit position of particles in the CBM detector planes, tracks reconstructed from the hits or combinations thereof. The Deep Learning models reconstruct impact parameters from 2-14 fm with a mean error varying from -0.33 to 0.22 fm. For impact parameters in the range of 5-14 fm, a model which uses the combination of hit and track information of particles has a relative precision of 4-9 % and a mean error of -0.33 to 0.13 fm. In the same range of impact parameters, a model with only track information has a relative precision of 4-10 % and a mean error of -0.18 to 0.22 fm. This new method of event-classification is shown to be more accurate and less model dependent than conventional methods and can utilize the performance boost of modern GPU processor units.

hep-ph↗

arXiv2020

Enhanced Bayesian Model Updating with Incomplete Modal Information Using Parallel, Interactive and Adaptive Markov Chains

Finite element model updating is challenging because 1) the problem is oftentimes underdetermined while the measurements are limited and/or incomplete; 2) many combinations of parameters may yield responses that are similar with respect to actual measurements; and 3) uncertainties inevitably exist. The aim of this research is to leverage upon computational intelligence through statistical inference to facilitate an enhanced, probabilistic finite element model updating using incomplete modal response measurement. This new framework is built upon efficient inverse identification through optimization, whereas Bayesian inference is employed to account for the effect of uncertainties. To overcome the computational cost barrier, we adopt Markov chain Monte Carlo (MCMC) to characterize the target function/distribution. Instead of using single Markov chain in conventional Bayesian approach, we develop a new sampling theory with multiple parallel, interactive and adaptive Markov chains and incorporate into Bayesian inference. This can harness the collective power of these Markov chains to realize the concurrent search of multiple local optima. The number of required Markov chains and their respective initial model parameters are automatically determined via Monte Carlo simulation-based sample pre-screening followed by K-means clustering analysis. These enhancements can effectively address the aforementioned challenges in finite element model updating. The validity of this framework is systematically demonstrated through case studies.

stat.AP↗