arXiv subjects

Zhi Wang

Publications and source records attributed to Zhi Wang.

At least 325 records · Page 18Linked to original sources

arXiv2019

AdaCompress: Adaptive Compression for Online Computer Vision Services

With the growth of computer vision based applications and services, an explosive amount of images have been uploaded to cloud servers which host such computer vision algorithms, usually in the form of deep learning models. JPEG has been used as the {\em de facto} compression and encapsulation method before one uploads the images, due to its wide adaptation. However, standard JPEG configuration does not always perform well for compressing images that are to be processed by a deep learning model, e.g., the standard quality level of JPEG leads to 50\% of size overhead (compared with the best quality level selection) on ImageNet under the same inference accuracy in popular computer vision models including InceptionNet, ResNet, etc. Knowing this, designing a better JPEG configuration for online computer vision services is still extremely challenging: 1) Cloud-based computer vision models are usually a black box to end-users; thus it is difficult to design JPEG configuration without knowing their model structures. 2) JPEG configuration has to change when different users use it. In this paper, we propose a reinforcement learning based JPEG configuration framework. In particular, we design an agent that adaptively chooses the compression level according to the input image's features and backend deep learning models. Then we train the agent in a reinforcement learning way to adapt it for different deep learning cloud services that act as the {\em interactive training environment} and feeding a reward with comprehensive consideration of accuracy and data size. In our real-world evaluation on Amazon Rekognition, Face++ and Baidu Vision, our approach can reduce the size of images by 1/2 -- 1/3 while the overall classification accuracy only decreases slightly.

cs.MM↗

arXiv2019

Decentralized Federated Learning: A Segmented Gossip Approach

The emerging concern about data privacy and security has motivated the proposal of federated learning, which allows nodes to only synchronize the locally-trained models instead their own original data. Conventional federated learning architecture, inherited from the parameter server design, relies on highly centralized topologies and the assumption of large nodes-to-server bandwidths. However, in real-world federated learning scenarios the network capacities between nodes are highly uniformly distributed and smaller than that in a datacenter. It is of great challenges for conventional federated learning approaches to efficiently utilize network capacities between nodes. In this paper, we propose a model segment level decentralized federated learning to tackle this problem. In particular, we propose a segmented gossip approach, which not only makes full utilization of node-to-node bandwidth, but also has good training convergence. The experimental results show that even the training time can be highly reduced as compared to centralized federated learning.

cs.LG↗

arXiv2019

An Exploratory Analysis of the Latent Structure of Process Data via Action Sequence Autoencoder

Computer simulations have become a popular tool of assessing complex skills such as problem-solving skills. Log files of computer-based items record the entire human-computer interactive processes for each respondent. The response processes are very diverse, noisy, and of nonstandard formats. Few generic methods have been developed for exploiting the information contained in process data. In this article, we propose a method to extract latent variables from process data. The method utilizes a sequence-to-sequence autoencoder to compress response processes into standard numerical vectors. It does not require prior knowledge of the specific items and human-computers interaction patterns. The proposed method is applied to both simulated and real process data to demonstrate that the resulting latent variables extract useful information from the response processes.

stat.ML↗

arXiv2019

Load Balancing for Ultra-Dense Networks: A Deep Reinforcement Learning Based Approach

In this paper, we propose a deep reinforcement learning (DRL) based mobility load balancing (MLB) algorithm along with a two-layer architecture to solve the large-scale load balancing problem for ultra-dense networks (UDNs). Our contribution is three-fold. First, this work proposes a two-layer architecture to solve the large-scale load balancing problem in a self-organized manner. The proposed architecture can alleviate the global traffic variations by dynamically grouping small cells into self-organized clusters according to their historical loads, and further adapt to local traffic variations through intra-cluster load balancing afterwards. Second, for the intra-cluster load balancing, this paper proposes an off-policy DRL-based MLB algorithm to autonomously learn the optimal MLB policy under an asynchronous parallel learning framework, without any prior knowledge assumed over the underlying UDN environments. Moreover, the algorithm enables joint exploration with multiple behavior policies, such that the traditional MLB methods can be used to guide the learning process thereby improving the learning efficiency and stability. Third, this work proposes an offline-evaluation based safeguard mechanism to ensure that the online system can always operate with the optimal and well-trained MLB policy, which not only stabilizes the online performance but also enables the exploration beyond current policies to make full use of machine learning in a safe way. Empirical results verify that the proposed framework outperforms the existing MLB methods in general UDN environments featured with irregular network topologies, coupled interferences, and random user movements, in terms of the load balancing performance.

cs.LG↗

arXiv2019

Realization of predicted exotic materials: The burden of proof

Trove of exotic topoloid structures has recently been predicted by searching for compounds whose calculated band structure crossing points fulfill specific symmetry requirements. Discovery of exciting physical phenomena by experimental studies of such predicted compounds is just around the corner. Yet, the examination of some of these assumed high-symmetry structures suggests that not always will assembly of atoms in a configuration that yields exotic topological properties be protected against energy-lowering symmetry breaking modes. Indeed, although bulk topological characteristics lead to protected surface/edge states, nothing protects bulk states from structural instability. The burden of proof for theoretical predictions of exciting physical phenomena should include some compelling hints that such phenomena can live in thermodynamically stable (or near stable) compounds. Herein, we illustrate how the use of the calculated total (electron + ion) energies of candidate structures can remove false-positive predicted topoloids from the list of likely realizable compounds, to the benefit of the much-cherished iterative process of theory-experiment materials discovery.

cond-mat.mtrl-sci↗

arXiv2019

Annotation-Free Cardiac Vessel Segmentation via Knowledge Transfer from Retinal Images

Segmenting coronary arteries is challenging, as classic unsupervised methods fail to produce satisfactory results and modern supervised learning (deep learning) requires manual annotation which is often time-consuming and can some time be infeasible. To solve this problem, we propose a knowledge transfer based shape-consistent generative adversarial network (SC-GAN), which is an annotation-free approach that uses the knowledge from publicly available annotated fundus dataset to segment coronary arteries. The proposed network is trained in an end-to-end fashion, generating and segmenting synthetic images that maintain the background of coronary angiography and preserve the vascular structures of retinal vessels and coronary arteries. We train and evaluate the proposed model on a dataset of 1092 digital subtraction angiography images, and experiments demonstrate the supreme accuracy of the proposed method on coronary arteries segmentation.

eess.IV↗

arXiv2019

A Pvalue-guided Anomaly Detection Approach Combining Multiple Heterogeneous Log Parser Algorithms on IIoT Systems

Industrial Internet of Things (IIoT) is becoming an attack target of advanced persistent threat (APT). Currently, IIoT logs have not been effectively used for anomaly detection. In this paper, we use blockchain to prevent logs from being tampered with and propose a pvalue-guided anomaly detection approach. This approach uses statistical pvalues to combine multiple heterogeneous log parser algorithms. The weighted edit distance is selected as a score function to calculate the nonconformity score between a log and a predefined event. The pvalue is calculated based on the non-conformity scores which indicate how well a log matches an event. This approach is tested on a large number of real-world HDFS logs and IIoT logs. The experiment results show that abnormal events could be effectively recognized by our pvalue-guided approach.

cs.CR↗

arXiv2019

Alloy theory with atomic resolution for Rashba or topological systems

Interest in substitutional disordered alloys has recently reemerged with focus on the symmetry-sensitive properties in the alloy such as topological insulation and Rashba effect. A substitutional random alloy manifests a distribution of local environments, creating a polymorphous network. While the macroscopic average (monomorphous) structure may have the original high symmetry of the constituent compounds, many observable physical properties are sensitive to local symmetry, and are hence $ $ rather than $P(S_0)$=$P( )$. The fundamental difference between polymorphous $ $ and monomorphous $P(S_0)$ led to the often-diverging results and the missing the atomic-scale resolution needed to discern symmetry-related physics. A natural approach capturing the polymorphous aspect is supercell model, which however suffers the difficulty of band folding ('spaghetti bands'), rendering the E vs k dispersion needed in topology and Rashba physics and seen in experiments, practically inaccessible. A solution that retains the polymorphous nature but restores the E vs k relation is to unfold the supercell bands. This yields alloy Effective Band Structure (EBS), providing a 3D picture of spectral density consisting of E- and k-dependent spectral weight with coherent and incoherent features, all created naturally by the polymorphous distribution of many local environments. We illustrate this EBS approach for CdTe-HgTe, PbSe-SnSe and PbS-PbTe alloys. We found properties that are critical for e.g. topological phase transition and Rashba splitting but totally absent in conventional monomorphous approaches, including (1) co-existing, wavevector- and energy-dependent coherent band splitting and incoherent band broadening, (2) coherent-incoherent transition along different k space directions, and (3) Rashba-like band splitting having both coherent and incoherent features.

cond-mat.mtrl-sci↗

arXiv2019

Latent Feature Extraction for Process Data via Multidimensional Scaling

Computer-based interactive items have become prevalent in recent educational assessments. In such items, the entire human-computer interactive process is recorded in a log file and is known as the response process. This paper aims at extracting useful information from response processes. In particular, we consider an exploratory latent variable analysis for process data. Latent variables are extracted through a multidimensional scaling framework and can be empirically proved to contain more information than classic binary responses in terms of out-of-sample prediction of many variables.

stat.AP↗

arXiv2019

The Wu-Yang potential of Magnetic Skyrmion from SU(2) Flat Connection

The theoretical research of the origin of magnetic skyrmion is very interesting. By using decomposition theory of gauge potential and the gauge parallel condition of local bases of $su(2)$ Lie algebra, its $SU(2)$ gauge potential is expressed as flat connection. As an example of application, we obtain the inner topological structure of second Chern number by $SU(2)$ flat connection method. It's well known that if magnetic monopole exists in electrodynamics, its Wu-Yang potential is indispensable in $U(1)$ invariant electromagnetic field. In $2$-dim magnetic materials, we prove that if magnetic skyrmion exists, its integral kernel must be $U(1)$ Wu-Yang curvature, where its $U(1)$ Wu-Yang potential is the projection of $SU(2)$ flat connection on $su(2)$ local Cartan subalgebra. The magnetic skyrmion can be created by performing concrete $SU(2)$ local gauge transformation to $su(2)$ Cartan subalgebra $σ_3$. The components of the $U(1)$ Wu-Yang curvature correspond to the emergent electromagnetic field of magnetic skyrmion

hep-th↗

arXiv2019

Disorder-Induced Weyl Semimetal Phase and Sequential Band Inversions in PbSe-SnSe Alloys

The search for topological systems has recently broadened to include random substitutional alloys, which lack the specific crystalline symmetries that protect topological phases, raising the question whether topological properties can be preserved, or are modified by disorder. To address this question, we avoid methods that assumed at the outset high (averaged) symmetry, using instead a fully-atomistic, topological description of alloy. Application to PbSe-SnSe alloy reveals that topology survives in an interesting fashion: (a) spatial randomness removes the valley degeneracy (splitting larger than 150 meV), leading to a sequential inversion of the split valley components over a range of compositions; (b) absence of inversion lifts spin degenerates, leading to a Weyl semimetal phase without the need of external magnetic field, an unexpected result, given that the alloy constituent compounds are inversion-symmetric. (a) and (b) underpin the topological physics at low symmetry and complete the missing understanding of possible topological phases within the normal-topological insulator transition.

cond-mat.mtrl-sci↗

arXiv2019

Towards Constraining Parity-Violations in Gravity with Satellite Gradiometry

Parity violation in gravity, if existed, could have important implications, and it is meaningful to search and test the possible observational effects. Chern-Simons modified gravity serves as a natural model for gravitational parity-violations. Especially, considering extensions to Einstein-Hilbert action up to second order curvature terms, it is known that theories of gravitational parity-violation will reduce to the dynamical Chern-Simons gravity. In this letter, we outline the theoretical principles of testing the dynamical Chern-Simons gravity with orbiting gravity gradiometers, which could be naturally incorporated into future satellite gravity missions. The secular gravity gradient signals, due to the Mashhoon-Theiss (anomaly) effect, in dynamical Chern-Simons gravity are worked out, which can improve the constraint of the corresponding Chern-Simons length scale $ξ^{\frac{1}{4}}_{cs}$ obtained from such measurement scheme. For orbiting superconducting gradiometers or gradiometers with optical readout, a bound $ξ^{\frac{1}{4}}_{cs}\leq 10^6 \ km$ (or even better) could in principle be obtained, which will be at least 2 orders of magnitude stronger than the current one based on the observations from the GP-B mission and the LAGEOS I, II satellites.

gr-qc↗

arXiv2018

JALAD: Joint Accuracy- and Latency-Aware Deep Structure Decoupling for Edge-Cloud Execution

Recent years have witnessed a rapid growth of deep-network based services and applications. A practical and critical problem thus has emerged: how to effectively deploy the deep neural network models such that they can be executed efficiently. Conventional cloud-based approaches usually run the deep models in data center servers, causing large latency because a significant amount of data has to be transferred from the edge of network to the data center. In this paper, we propose JALAD, a joint accuracy- and latency-aware execution framework, which decouples a deep neural network so that a part of it will run at edge devices and the other part inside the conventional cloud, while only a minimum amount of data has to be transferred between them. Though the idea seems straightforward, we are facing challenges including i) how to find the best partition of a deep structure; ii) how to deploy the component at an edge device that only has limited computation power; and iii) how to minimize the overall execution latency. Our answers to these questions are a set of strategies in JALAD, including 1) A normalization based in-layer data compression strategy by jointly considering compression rate and model accuracy; 2) A latency-aware deep decoupling strategy to minimize the overall execution latency; and 3) An edge-cloud structure adaptation strategy that dynamically changes the decoupling for different network conditions. Experiments demonstrate that our solution can significantly reduce the execution latency: it speeds up the overall inference execution with a guaranteed model accuracy loss.

cs.DC↗

arXiv2018

Phonon-assisted Auger enables ultrafast charge transfer in CdSe Quantum Dot/Organic Molecule

Charge transfer between photoexcited quantum dots and molecular acceptors is one of the key limiting processes in most applications of colloidal nanostructures, most prominently in photovoltaics. An atomistic detailed description of this process would open new ways to optimize existing and create new structures with targeted properties. We achieve a one-to-one comparison between ab-initio non-adiabatic molecular dynamics calculations and transient absorption spectroscopy experiments, which allows us to draw a comprehensive atomistic picture of the charge transfer process, following the time evolution of the charge carrier across the electronic landscape and identifying the thereby induced vibrations. For two quantum dot sizes we find two qualitatively different processes. For the larger structure we find a relatively slow (τ= 516 fs) transfer process that we explain by the existence of a large energy detuning and weak vibronic coupling. For the smaller structure the process is ultrafast (τ= 20 fs) due to an efficient, phonon-assisted Auger process triggered by a strong electron-hole coupling.

cond-mat.mtrl-sci↗

arXiv2018

Coherence-Based Performance Guarantee of Regularized $\ell_{1}$-Norm Minimization and Beyond

In this paper, we consider recovering the signal $\bm{x}\in\mathbb{R}^{n}$ from its few noisy measurements $\bm{b}=A\bm{x}+\bm{z}$, where $A\in\mathbb{R}^{m\times n}$ with $m\ll n$ is the measurement matrix, and $\bm{z}\in\mathbb{R}^{m}$ is the measurement noise/error. We first establish a coherence-based performance guarantee for a regularized $\ell_{1}$-norm minimization model to recover such signals $\bm{x}$ in the presence of the $\ell_{2}$-norm bounded noise, i.e., $\|\bm{z}\|_{2}\leqε$, and then extend these theoretical results to guarantee the robust recovery of the signals corrupted with the Dantzig Selector (DS) type noise, i.e., $\|A^{T}\bm{z}\|_{\infty}\leqε$, and the structured block-sparse signal recovery in the presence of the bounded noise. To the best of our knowledge, we first extend nontrivially the sharp uniform recovery condition derived by Cai, Wang and Xu (2010) for the constrained $\ell_{1}$-norm minimization model, which takes the form of \begin{align*} μ<\frac{1}{2k-1}, \end{align*} where $μ$ is defined as the (mutual) coherence of $A$, to two unconstrained regularized $\ell_{1}$-norm minimization models to guarantee the robust recovery of any signals (not necessary to be $k$-sparse) under the $\ell_{2}$-norm bounded noise and the DS type noise settings, respectively. Besides, a uniform recovery condition and its two resulting error estimates are also established for the first time to our knowledge, for the robust block-sparse signal recovery using a regularized mixed $\ell_{2}/\ell_{1}$-norm minimization model, and these results well complement the existing theoretical investigation on this model which focuses on the non-uniform recovery conditions and/or the robust signal recovery in presence of the random noise.

math.NA↗

arXiv2018

Hysteresis from nonlinear dynamics of Majorana modes in topological Josephson junctions

We reveal that topological Josephson junctions provide a natural platform for the interplay between the Josephson effect and the Landau-Zener effect through a two-level system formed by coupled Majorana modes. We build a quantum resistively shunted junction (RSJ) model by modifying the standard textbook RSJ model to take account of the two-level system from the Majorana modes at the junction. We show that the dynamics of the two-level system is governed by a nonlinear Schrödinger equation and solve the equations analytically via a mapping to a classical dynamical problem. This nonlinear dynamics leads to hysteresis in the I-V characteristics, which can give a quantitative explanation to recent experiments. We also predict the coexistence of two interference patterns with periods $h/e$ and $h/2e$ in topological superconducting quantum interference devices.

cond-mat.supr-con↗

arXiv2018

3D characterization of ultrasonic melt processing on the microstructural refinement of Al-Cu alloys by synchrotron X-ray tomography

The effect of ultrasonic melting processing on three-dimensional architecture of intermetallic phases and pores in two multicomponent cast Al-5.0Cu-0.6Mn-0.5 Fe alloys is characterized using conventional microscopy and synchrotron X-ray microtomography. The two alloys are found to contain intermetallic phases such as Al15(FeMn)3Cu2, Al7Cu2Fe, Al3(FeMn), Al6(FeMn), and Al2Cu that have complex networked morphology in 3D. The application of USP in alloys can obtained refined and equiaxed microstructures. The grain size of 0.5Fe and 1.0 Fe alloys is greatly decreased from 16.9 m, 15.8 m without USP to 13.3 m, 12.2 m with USP, respectively. The results show that USP significantly reduce the volume fraction, grain size, interconnectivity, and equivalent diameter of the intermetallic phases in both alloys. The volume fraction of pores in both alloys is reduced due to the USP degassing effect. The refinement mechanism of USP induced fragmentation of primary and secondary dendrites via acoustic bubbles and acoustic streaming flow were discussed.

cond-mat.mtrl-sci↗

arXiv2018

Handover Control in Wireless Systems via Asynchronous Multi-User Deep Reinforcement Learning

In this paper, we propose a two-layer framework to learn the optimal handover (HO) controllers in possibly large-scale wireless systems supporting mobile Internet-of-Things (IoT) users or traditional cellular users, where the user mobility patterns could be heterogeneous. In particular, our proposed framework first partitions the user equipments (UEs) with different mobility patterns into clusters, where the mobility patterns are similar in the same cluster. Then, within each cluster, an asynchronous multi-user deep reinforcement learning scheme is developed to control the HO processes across the UEs in each cluster, in the goal of lowering the HO rate while ensuring certain system throughput. In this scheme, we use a deep neural network (DNN) as an HO controller learned by each UE via reinforcement learning in a collaborative fashion. Moreover, we use supervised learning in initializing the DNN controller before the execution of reinforcement learning to exploit what we already know with traditional HO schemes and to mitigate the negative effects of random exploration at the initial stage. Furthermore, we show that the adopted global-parameter-based asynchronous framework enables us to train faster with more UEs, which could nicely address the scalability issue to support large systems. Finally, simulation results demonstrate that the proposed framework can achieve better performance than the state-of-art on-line schemes, in terms of HO rates.

cs.NI↗