arXiv subjects

Zhi Wang

Publications and source records attributed to Zhi Wang.

At least 289 records · Page 16Linked to original sources

arXiv2021

Multitask Bandit Learning Through Heterogeneous Feedback Aggregation

In many real-world applications, multiple agents seek to learn how to perform highly related yet slightly different tasks in an online bandit learning protocol. We formulate this problem as the $ε$-multi-player multi-armed bandit problem, in which a set of players concurrently interact with a set of arms, and for each arm, the reward distributions for all players are similar but not necessarily identical. We develop an upper confidence bound-based algorithm, RobustAgg$(ε)$, that adaptively aggregates rewards collected by different players. In the setting where an upper bound on the pairwise similarities of reward distributions between players is known, we achieve instance-dependent regret guarantees that depend on the amenability of information sharing across players. We complement these upper bounds with nearly matching lower bounds. In the setting where pairwise similarities are unknown, we provide a lower bound, as well as an algorithm that trades off minimax regret guarantees for adaptivity to unknown similarity structure.

cs.LG↗

arXiv2021

Towards Efficient Compressive Data Collection in the Internet of Things

It is of paramount importance to achieve efficient data collection in the Internet of Things (IoT). Due to the inherent structural properties (e.g., sparsity) existing in many signals of interest, compressive sensing (CS) technology has been extensively used for data collection in IoT to improve both accuracy and energy efficiency. Apart from the existing works which leverage CS as a channel coding scheme to deal with data loss during transmission, some recent results have started to employ CS as a source coding strategy. The frequently used projection matrices in these CS-based source coding schemes include dense random matrices (e.g., Gaussian matrices or Bernoulli matrices) and structured matrices (e.g., Toeplitz matrices). However, these matrices are either difficult to be implemented on resource-constrained IoT sensor nodes or have limited applicability. To address these issues, in this paper, we design a novel simple and efficient projection matrix, named sparse Gaussian matrix, which is easy and resource-saving to be implemented in practical IoT applications. We conduct both theoretical analysis and experimental evaluation of the designed sparse Gaussian matrix. The results demonstrate that employing the designed projection matrix to perform CS-based source coding could significantly save time and memory cost while ensuring satisfactory signal recovery performance.

cs.IT↗

arXiv2021

Multimedia Edge Computing

In this paper, we investigate the recent studies on multimedia edge computing, from sensing not only traditional visual/audio data but also individuals' geographical preference and mobility behaviors, to performing distributed machine learning over such data using the joint edge and cloud infrastructure and using evolutional strategies like reinforcement learning and online learning at edge devices to optimize the quality of experience for multimedia services at the last mile proactively. We provide both a retrospective view of recent rapid migration (resp. merge) of cloud multimedia to (resp. and) edge-aware multimedia and insights on the fundamental guidelines for designing multimedia edge computing strategies that target satisfying the changing demand of quality of experience. By showing the recent research studies and industrial solutions, we also provide future directions towards high-quality multimedia services over edge computing.

cs.MM↗

arXiv2021

Extract then Distill: Efficient and Effective Task-Agnostic BERT Distillation

Task-agnostic knowledge distillation, a teacher-student framework, has been proved effective for BERT compression. Although achieving promising results on NLP tasks, it requires enormous computational resources. In this paper, we propose Extract Then Distill (ETD), a generic and flexible strategy to reuse the teacher's parameters for efficient and effective task-agnostic distillation, which can be applied to students of any size. Specifically, we introduce two variants of ETD, ETD-Rand and ETD-Impt, which extract the teacher's parameters in a random manner and by following an importance metric respectively. In this way, the student has already acquired some knowledge at the beginning of the distillation process, which makes the distillation process converge faster. We demonstrate the effectiveness of ETD on the GLUE benchmark and SQuAD. The experimental results show that: (1) compared with the baseline without an ETD strategy, ETD can save 70\% of computation cost. Moreover, it achieves better results than the baseline when using the same computing resource. (2) ETD is generic and has been proven effective for different distillation methods (e.g., TinyBERT and MiniLM) and students of different sizes. The source code will be publicly available upon publication.

cs.CL↗

arXiv2021

Weak cosmic censorship conjecture is not violated for a rotating linear dilaton black hole

In this paper, we investigate the validity of the weak cosmic censorship conjecture(WCCC) for a rotating linear dilaton black hole from two different methods. By using the classsical ingoing test particle method, we obtain the same results as given in the new version of gedanken experiment recently proposed by Wald. We find that even for this rotating linear dilaton black hole the Iyer-Wald formalism is still functioning. By comparing these two methods, we find that the same result will be obtained in both cases up to first order. The nearly extremal black hole can be overspun, while the extremal one cannot be overspun. When we include the second order modification into consideration, the Iyer-Wald method show that even for the nearly extremal black hole, the WCCC is well protected. These results imply that weak cosmic censorship conjecture is still valid using the ingoing test particle method up to second order modification.

gr-qc↗

arXiv2021

Rule-Based Reinforcement Learning for Efficient Robot Navigation with Space Reduction

For real-world deployments, it is critical to allow robots to navigate in complex environments autonomously. Traditional methods usually maintain an internal map of the environment, and then design several simple rules, in conjunction with a localization and planning approach, to navigate through the internal map. These approaches often involve a variety of assumptions and prior knowledge. In contrast, recent reinforcement learning (RL) methods can provide a model-free, self-learning mechanism as the robot interacts with an initially unknown environment, but are expensive to deploy in real-world scenarios due to inefficient exploration. In this paper, we focus on efficient navigation with the RL technique and combine the advantages of these two kinds of methods into a rule-based RL (RuRL) algorithm for reducing the sample complexity and cost of time. First, we use the rule of wall-following to generate a closed-loop trajectory. Second, we employ a reduction rule to shrink the trajectory, which in turn effectively reduces the redundant exploration space. Besides, we give the detailed theoretical guarantee that the optimal navigation path is still in the reduced space. Third, in the reduced space, we utilize the Pledge rule to guide the exploration strategy for accelerating the RL process at the early stage. Experiments conducted on real robot navigation problems in hex-grid environments demonstrate that RuRL can achieve improved navigation performance.

cs.RO↗

arXiv2021

Measuring Discrimination to Boost Comparative Testing for Multiple Deep Learning Models

The boom of DL technology leads to massive DL models built and shared, which facilitates the acquisition and reuse of DL models. For a given task, we encounter multiple DL models available with the same functionality, which are considered as candidates to achieve this task. Testers are expected to compare multiple DL models and select the more suitable ones w.r.t. the whole testing context. Due to the limitation of labeling effort, testers aim to select an efficient subset of samples to make an as precise rank estimation as possible for these models. To tackle this problem, we propose Sample Discrimination based Selection (SDS) to select efficient samples that could discriminate multiple models, i.e., the prediction behaviors (right/wrong) of these samples would be helpful to indicate the trend of model performance. To evaluate SDS, we conduct an extensive empirical study with three widely-used image datasets and 80 real world DL models. The experimental results show that, compared with state-of-the-art baseline methods, SDS is an effective and efficient sample selection method to rank multiple DL models.

cs.SE↗

arXiv2021

Strong influence of non-magnetic ligands on the momentum dependent spin splitting in antiferromagnets

Recent studies have shown that the non-relativistic antiferromagnetic ordering could generate momentum-dependent spin splitting analogous to the Rashba effect, but free from the requirement of relativistic spin-orbit coupling. Whereas the classification of such compounds can be illustrated by different spin-splitting prototypes (SSTs) from symmetry analysis and density functional theory calculations, the significant variation in bonding and structure of these diverse compounds representing different SSTs clouds the issue of how much of the variation in spin splitting can be traced back to the symmetry-defined characteristics, rather to the underlining chemical and structural diversity. The alternative model Hamiltonian approaches do not confront the issues of chemical and structural complexity, but often consider only the magnetic sublattice, dealing with the all-important effects of the non-magnetic ligands via renormalizing the interactions between the magnetic sites. To this end we constructed a 'DFT model Hamiltonian' that allows us to study SSTs at approximate 'constant chemistry', while retaining the realistic atomic scale structure including ligands. This is accomplished by using a single, universal magnetic skeletal lattice (Ni2+ ions in Rocksalt NiO) and designing small displacements of the non-magnetic (oxygen) sublattice which produce, by design, the different SSTs magnetic symmetries. We show that (i) even similar crystal structures having very similar band structures can lead to contrasting behavior of spin splitting vs. momentum, and (ii) even subtle deformations of the non-magnetic ligand sublattice could cause a giant spin splitting in AFM-induced SST. This is a paradigm shift relative to the convention of modeling magnets without considering the non-magnetic ligand that mediate indirect magnetic interaction (e.g., super exchange).

cond-mat.mtrl-sci↗

arXiv2021

Tinker-HP : Accelerating Molecular Dynamics Simulations of Large Complex Systems with Advanced Point Dipole Polarizable Force Fields using GPUs and Multi-GPUs systems

We present the extension of the Tinker-HP package (Lagardère et al., Chem. Sci., 2018,9, 956-972) to the use of Graphics Processing Unit (GPU) cards to accelerate molecular dynamics simulations using polarizable many-body force fields. The new high-performance module allows for an efficient use of single- and multi-GPU architectures ranging from research laboratories to modern supercomputer centers. After detailing an analysis of our general scalable strategy that relies on OpenACC and CUDA, we discuss the various capabilities of the package. Among them, the multi-precision possibilities of the code are discussed. If an efficient double precision implementation is provided to preserve the possibility of fast reference computations, we show that a lower precision arithmetic is preferred providing a similar accuracy for molecular dynamics while exhibiting superior performances. As Tinker-HP is mainly dedicated to accelerate simulations using new generation point dipole polarizable force field, we focus our study on the implementation of the AMOEBA model. Testing various NVIDIA platforms including 2080Ti, 3090, V100 and A100 cards, we provide illustrative benchmarks of the code for single- and multi-cards simulations on large biosystems encompassing up to millions of atoms. The new code strongly reduces time to solution and offers the best performances to date obtained using the AMOEBA polarizable force field. Perspectives toward the strong-scaling performance of our multi-node massive parallelization strategy, unsupervised adaptive sampling and large scale applicability of the Tinker-HP code in biophysics are discussed. The present software has been released in phase advance on GitHub in link with the High Performance Computing community COVID-19 research efforts and is free for Academics (see https://github.com/TinkerTools/tinker-hp).

physics.comp-ph↗

arXiv2021

Towards Understanding and Demystifying Bitcoin Mixing Services

One reason for the popularity of Bitcoin is due to its anonymity. Although several heuristics have been used to break the anonymity, new approaches are proposed to enhance its anonymity at the same time. One of them is the mixing service. Unfortunately, mixing services have been abused to facilitate criminal activities, e.g., money laundering. As such, there is an urgent need to systematically understand Bitcoin mixing services. In this paper, we take the first step to understand state-of-the-art Bitcoin mixing services. Specifically, we propose a generic abstraction model for mixing services and observe that there are two mixing mechanisms in the wild, i.e. {swapping} and {obfuscating}. Based on this model, we conduct a transaction-based analysis and successfully reveal the mixing mechanisms of four representative services. Besides, we propose a method to identify mixing transactions that leverage the obfuscating mechanism. The proposed approach is able to identify over $92$\% of the mixing transactions. Based on identified transactions, we then estimate the profit of mixing services and provide a case study of tracing the money flow of stolen Bitcoins.

cs.CR↗

arXiv2021

Lifelong Incremental Reinforcement Learning with Online Bayesian Inference

A central capability of a long-lived reinforcement learning (RL) agent is to incrementally adapt its behavior as its environment changes, and to incrementally build upon previous experiences to facilitate future learning in real-world scenarios. In this paper, we propose LifeLong Incremental Reinforcement Learning (LLIRL), a new incremental algorithm for efficient lifelong adaptation to dynamic environments. We develop and maintain a library that contains an infinite mixture of parameterized environment models, which is equivalent to clustering environment parameters in a latent space. The prior distribution over the mixture is formulated as a Chinese restaurant process (CRP), which incrementally instantiates new environment models without any external information to signal environmental changes in advance. During lifelong learning, we employ the expectation maximization (EM) algorithm with online Bayesian inference to update the mixture in a fully incremental manner. In EM, the E-step involves estimating the posterior expectation of environment-to-cluster assignments, while the M-step updates the environment parameters for future learning. This method allows for all environment models to be adapted as necessary, with new models instantiated for environmental changes and old models retrieved when previously seen environments are encountered again. Experiments demonstrate that LLIRL outperforms relevant existing methods, and enables effective incremental adaptation to various dynamic environments for lifelong learning.

cs.LG↗

arXiv2020

Thermal Transport in Polymers: A Review

In this article, we review thermal transport in polymers with different morphologies from aligned fibers to bulk amorphous states. We survey early and recent efforts in engineering polymers with high thermal conductivity by fabricating polymers with large-scale molecular alignments. The experimentally realized extremely high thermal conductivity of polymer nanofibers are highlighted, and understanding of thermal transport physics from molecular simulations are discussed. We then transition to the discussion of bulk amorphous polymers with an emphasize on the physics of thermal transport and its relation with the conformation of molecular chains in polymers. We also discuss the current understanding of how the chemistry of polymers would influence thermal transport in amorphous polymers and some limited, but important chemistry-structural-property relationships. Lastly, challenges, perspectives and outlook of this field are presented. We hope this review will inspire more fundamental and applied research in the polymer thermal transport field to advance scientific understanding and engineering applications.

cond-mat.mtrl-sci↗

arXiv2020

STSIR: Spatial Temporal Pandemic Model with Mobility Data

With the outbreak of COVID-19, how to mitigate and suppress its spread is a big issue to the government. Department of public health need powerful models to model and predict the trend and scale of such pandemic. And models that could evaluate the effect of the public policy are also essential to the fight with the COVID-19. A main limitation of existing models is that they can only evaluate the policy by calculating $R_0$ after infection happens instead of giving observable index. To tackle this, based on the transmission character of the COVID-19, we preposed a novel framework Spatial-Temporal-Susceptible-Infected-Removed (STSIR) model. In particular, we merged both intra-city and inter-city mobility index with the traditional SIR dynamics and make it a dynamic system. And we proved that the STSIR system is a closed system which makes the system self-consistent. And finally we proposed a Multi-Stage Simulated Annealing (MSSA) algorithm to find optimal parameter of the system. In our experiments, based on Baidu Mobility dataset, and China pandemic dataset provided by Dingxiangyuan, our model can effectively predict the total scale of the pandemic and also gives clear policy analysis with observable index.

cs.CE↗

arXiv2020

CryptMPI: A Fast Encrypted MPI Library

The cloud infrastructure must provide security for High-Performance Computing (HPC) applications of sensitive data to execute in such an environment. However, supporting security in the communication infrastructure of today's public cloud is challenging, because current networks for data centers are so fast that adding encryption can incur very significant overheads. In this work, we introduce CryptMPI, a high performance encrypted MPI library that supports communication with both integrity and privacy. We present the techniques in CryptMPI and report our benchmarking results using micro-benchmarks and NAS parallel benchmarks. The evaluation results indicate that the aforementioned techniques are effective in improving the performance of encrypted communication.

cs.DC↗

arXiv2020

Time-Travel Investigation: Towards Building A Scalable Attack Detection Framework on Ethereum

As one of the representative blockchain platforms, Ethereum has attracted lots of attacks. Due to the existed financial loss, there is a pressing need to perform timely investigation and detect more attack instances. Though multiple systems have been proposed, they suffer from the scalability issue due to the following reasons. First, the tight coupling between malicious contract detection and blockchain data importing makes them infeasible to repeatedly detect different attacks. Second, the coarse-grained archive data makes them inefficient to replay transactions. Third, the separation between malicious contract detection and runtime state recovery consumes lots of storage. In this paper, we present the design of a scalable attack detection framework on Ethereum. It overcomes the scalability issue by saving the Ethereum state into a database and providing an efficient way to locate suspicious transactions. The saved state is fine-grained to support the replay of arbitrary transactions. The state is well-designed to avoid saving unnecessary state to optimize the storage consumption. We implement a prototype named EthScope and solve three technical challenges, i.e., incomplete Ethereum state, scalability, and extensibility. The performance evaluation shows that our system can solve the scalability issue, i.e., efficiently performing a large-scale analysis on billions of transactions, and a speedup of around 2,300x when replaying transactions. It also has lower storage consumption compared with existing systems. The result with three different types of information as inputs shows that our system can help an analyst understand attack behaviors and further detect more attacks. To engage the community, we will release our system and the dataset of detected attacks.

cs.CR↗

arXiv2020

Towards Efficiently Establishing Mutual Distrust Between Host Application and Enclave for SGX

Since its debut, SGX has been used in many applications, e.g., secure data processing. However, previous systems usually assume a trusted enclave and ignore the security issues caused by an untrusted enclave. For instance, a vulnerable (or even malicious) third-party enclave can be exploited to attack the host application and the rest of the system. In this paper, we propose an efficient mechanism to confine an untrusted enclave's behaviors. The threats of an untrusted enclave come from the enclave-host asymmetries. They can be abused to access arbitrary memory regions of its host application, jump to any code location after leaving the enclave and forge the stack register to manipulate the saved context. Our solution breaks such asymmetries and establishes mutual distrust between the host application and the enclave. It leverages Intel MPK for efficient memory isolation and the x86 single-step debugging mechanism to capture the event when an enclave is existing. It then performs the integrity check for the jump target and the stack pointer. We have solved two practical challenges and implemented a prototype system. The evaluation with multiple micro-benchmarks and representative real-world applications demonstrated the efficiency of our system, with less than 4% performance overhead.

cs.CR↗

arXiv2020

Understanding Electronic Peculiarities in Tetragonal FeSe as Local Structural Symmetry Breaking

Traditional band theory of perfect crystalline solids often uses as input the structure deduced from diffraction experiments; when modeled by the minimal unit cell this often produces a spatially averaged model. The present study illustrates that this is not always a safe practice unless one examines if the intrinsic bonding mechanism is capable of benefiting from the formation of a distribution of lower symmetry local environments that differ from the macroscopically averaged structure. This can happen either due to positional, or due to magnetic symmetry breaking. By removing the constraint of a small crystallographic cell, the energy minimization in the density functional theory finds atomic and spin symmetry breaking, not evident in conventional diffraction experiments but being found by local probes such as pair distribution function analysis. Here we report that large atomic and electronic anomalies in bulk tetragonal FeSe emerge from the existence of distributions of local positional and magnetic moment motifs. The found symmetry-broken motifs obtained by minimization of the internal energy represent what chemical bonding in tetragonal phase prefers as an intrinsic energy lowering static distortions. This explains observations of band renormalization, predicts orbital order and enhanced nematicity, and provides unprecedented close agreement with spectral function measured by photoemission and local atomic environment revealed by pair distribution function. While the symmetry-restricted strong correlation approach has been argued previously to be the exclusive theory needed for describing the main peculiarities of FeSe, we show here that the symmetry-broken mean-field approach addresses numerous aspects of the problem, provides intuitive insight into the electronic structure, and opens the door for large-scale mean-field calculations for similar d-electron quantum materials.

cond-mat.mtrl-sci↗

arXiv2020

Performance Evaluation and Modeling of Cryptographic Libraries for MPI Communications

In order for High-Performance Computing (HPC) applications with data security requirements to execute in the public cloud, the cloud infrastructure must ensure the privacy and integrity of data. To meet this goal, we consider incorporating encryption in the Message Passing Interface (MPI) library. We empirically evaluate four contemporary cryptographic libraries, OpenSSL, BoringSSL, Libsodium, and CryptoPP using micro-benchmarks and NAS parallel benchmarks on two different networking technologies, 10Gbps Ethernet and 40Gbps InfiniBand. We also develop accurate models that allow us to reason about the performance of encrypted MPI communication in different situations and give guidance on how to improve encrypted MPI performance.

cs.DC↗