Searcharxiv⌕ Search

arXiv subjects

Xiao Zhang

Publications and source records attributed to Xiao Zhang.

At least 451 records · Page 25Linked to original sources
arXiv2020

Learning Conceptual-Contextual Embeddings for Medical Text

External knowledge is often useful for natural language understanding tasks. We introduce a contextual text representation model called Conceptual-Contextual (CC) embeddings, which incorporates structured knowledge into text representations. Unlike entity embedding methods, our approach encodes a knowledge graph into a context model. CC embeddings can be easily reused for a wide range of tasks just like pre-trained language models. Our model effectively encodes the huge UMLS database by leveraging semantic generalizability. Experiments on electronic health records (EHRs) and medical text processing benchmarks showed our model gives a major boost to the performance of supervised medical NLP tasks.

cs.CL↗
arXiv2020

Predicting quantum many-body dynamics with transferable neural networks

Machine learning (ML) architectures such as convolutional neural networks (CNNs) have garnered considerable recent attention in the study of quantum many-body systems. However, advanced ML approaches such as transfer learning have seldom been applied to such contexts. Here we demonstrate that a simple recurrent unit (SRU) based efficient and transferable sequence learning framework is capable of learning and accurately predicting the time evolution of one-dimensional (1D) Ising model with simultaneous transverse and parallel magnetic fields, as quantitatively corroborated by relative entropy measurements and magnetization between the predicted and exact state distributions. At a cost of constant computational complexity, a larger many-body state evolution was predicted in an autoregressive way from just one initial state, without any guidance or knowledge of any Hamiltonian. Our work paves the way for future applications of advanced ML methods in quantum many-body dynamics only with knowledge from a smaller system.

cond-mat.stat-mech↗
arXiv2020

Understanding the Intrinsic Robustness of Image Distributions using Conditional Generative Models

Starting with Gilmer et al. (2018), several works have demonstrated the inevitability of adversarial examples based on different assumptions about the underlying input probability space. It remains unclear, however, whether these results apply to natural image distributions. In this work, we assume the underlying data distribution is captured by some conditional generative model, and prove intrinsic robustness bounds for a general class of classifiers, which solves an open problem in Fawzi et al. (2018). Building upon the state-of-the-art conditional generative models, we study the intrinsic robustness of two common image benchmarks under $\ell_2$ perturbations, and show the existence of a large gap between the robustness limits implied by our theory and the adversarial robustness achieved by current state-of-the-art robust models. Code for all our experiments is available at https://github.com/xiaozhanguva/Intrinsic-Rob.

cs.LG↗
arXiv2020

ACE -- An Anomaly Contribution Explainer for Cyber-Security Applications

In this paper, we introduce Anomaly Contribution Explainer or ACE, a tool to explain security anomaly detection models in terms of the model features through a regression framework, and its variant, ACE-KL, which highlights the important anomaly contributors. ACE and ACE-KL provide insights in diagnosing which attributes significantly contribute to an anomaly by building a specialized linear model to locally approximate the anomaly score that a black-box model generates. We conducted experiments with these anomaly detection models to detect security anomalies on both synthetic data and real data. In particular, we evaluate performance on three public data sets: CERT insider threat, netflow logs, and Android malware. The experimental results are encouraging: our methods consistently identify the correct contributing feature in the synthetic data where ground truth is available; similarly, for real data sets, our methods point a security analyst in the direction of the underlying causes of an anomaly, including in one case leading to the discovery of previously overlooked network scanning activity. We have made our source code publicly available.

cs.LG↗
arXiv2020

Label-guided Learning for Text Classification

Text classification is one of the most important and fundamental tasks in natural language processing. Performance of this task mainly dependents on text representation learning. Currently, most existing learning frameworks mainly focus on encoding local contextual information between words. These methods always neglect to exploit global clues, such as label information, for encoding text information. In this study, we propose a label-guided learning framework LguidedLearn for text representation and classification. Our method is novel but simple that we only insert a label-guided encoding layer into the commonly used text representation learning schemas. That label-guided layer performs label-based attentive encoding to map the universal text embedding (encoded by a contextual information learner) into different label spaces, resulting in label-wise embeddings. In our proposed framework, the label-guided layer can be easily and directly applied with a contextual encoding method to perform jointly learning. Text information is encoded based on both the local contextual information and the global label clues. Therefore, the obtained text embeddings are more robust and discriminative for text classification. Extensive experiments are conducted on benchmark datasets to illustrate the effectiveness of our proposed method.

cs.CL↗
arXiv2020

Tensor Graph Convolutional Networks for Text Classification

Compared to sequential learning models, graph-based neural networks exhibit some excellent properties, such as ability capturing global information. In this paper, we investigate graph-based neural networks for text classification problem. A new framework TensorGCN (tensor graph convolutional networks), is presented for this task. A text graph tensor is firstly constructed to describe semantic, syntactic, and sequential contextual information. Then, two kinds of propagation learning perform on the text graph tensor. The first is intra-graph propagation used for aggregating information from neighborhood nodes in a single graph. The second is inter-graph propagation used for harmonizing heterogeneous information between graphs. Extensive experiments are conducted on benchmark datasets, and the results illustrate the effectiveness of our proposed framework. Our proposed TensorGCN presents an effective way to harmonize and integrate heterogeneous information from different kinds of graphs.

cs.CL↗
arXiv2019

Multiplicity of stable orbits for deformable prolate capsules in shear flow

This work investigates the orbital dynamics of a fluid-filled deformable prolate capsule in unbounded shear flow. The motion of the capsule is simulated using a model that incorporates shear elasticity, area dilatation, and bending resistance. Here the deformability of the capsule is characterized by capillary number Ca, which represents the ratio of viscous stresses to elastic restoring stresses. For a capsule with small bending stiffness, at a given Ca, the orientation converges over time towards a unique stable orbit independent of the initial orientation. With increasing Ca, four dynamical modes are found for the stable orbit: rolling, wobbling, oscillating-swinging, and swinging. For a capsule with large bending stiffness, multiplicity in the orbit dynamics is observed. When the viscosity ratio $λ\lesssim 1$, the long-axis of the capsule always tends towards a stable orbit in the flow-gradient plane, either tumbling or swinging, depending on Ca. When $λ\gtrsim 1$, the stable orbit of the capsule is a tumbling motion at low Ca, irrespective of the initial orientation. Upon increasing Ca, there is a symmetry-breaking bifurcation away from the tumbling orbit, and the capsule is observed to adopt multiple stable orbital modes including nonsymmetric precessing and rolling, depending on the initial orientation. As Ca further increases, the nonsymmetric stable orbit loses existence at a saddle-node bifurcation, and rolling becomes the only attractor at high Ca, whereas the rolling state coexists with the nonsymmetric state at intermediate values of Ca. A symmetry-breaking bifurcation away from the rolling orbit is also found upon decreasing Ca. The regime with multiple attractors becomes broader as the aspect ratio of the capsule increases, while narrowing as viscosity ratio increases. We also report the particle contribution to the stress, which also displays multiplicity.

physics.flu-dyn↗
arXiv2019

Active Learning for Black-Box Adversarial Attacks in EEG-Based Brain-Computer Interfaces

Deep learning has made significant breakthroughs in many fields, including electroencephalogram (EEG) based brain-computer interfaces (BCIs). However, deep learning models are vulnerable to adversarial attacks, in which deliberately designed small perturbations are added to the benign input samples to fool the deep learning model and degrade its performance. This paper considers transferability-based black-box attacks, where the attacker trains a substitute model to approximate the target model, and then generates adversarial examples from the substitute model to attack the target model. Learning a good substitute model is critical to the success of these attacks, but it requires a large number of queries to the target model. We propose a novel framework which uses query synthesis based active learning to improve the query efficiency in training the substitute model. Experiments on three convolutional neural network (CNN) classifiers and three EEG datasets demonstrated that our method can improve the attack success rate with the same number of queries, or, in other words, our method requires fewer queries to achieve a desired attack performance. To our knowledge, this is the first work that integrates active learning and adversarial attacks for EEG-based BCIs.

cs.LG↗
arXiv2019

In-plane antiferromagnetic moments in axion topological insulator candidate EuIn$_2$As$_2$

Topological insulator with antiferromagnetic order can serve as an ideal platform for the realization of axion electrodynamics. In this paper, we report a systematic study of the axion topological insulator candidate EuIn$_2$As$_2$. A linear energy dispersion across the Fermi level confirms the existence of the proposed hole-type Fermi pocket. Spin-flop transitions occur with magnetic fields applied within the $ab$-plane while are absent for fields parallel to the $c$-axis. Anisotropic magnetic phase diagrams are observed and the orientation of the ground magnetic moment is found to be within the $ab$-plane. The magnetoresistivity for EuIn$_2$As$_2$ behaves non-monotonic as a function of field strength. It exhibits angular dependent evolving due to field-driven and temperature-driven magnetic states. These results indicate that the magnetic states of EuIn$_2$As$_2$ strongly affect the transport properties as well as the topological nature.

cond-mat.str-el↗
arXiv2019

SegSort: Segmentation by Discriminative Sorting of Segments

Almost all existing deep learning approaches for semantic segmentation tackle this task as a pixel-wise classification problem. Yet humans understand a scene not in terms of pixels, but by decomposing it into perceptual groups and structures that are the basic building blocks of recognition. This motivates us to propose an end-to-end pixel-wise metric learning approach that mimics this process. In our approach, the optimal visual representation determines the right segmentation within individual images and associates segments with the same semantic classes across images. The core visual learning problem is therefore to maximize the similarity within segments and minimize the similarity between segments. Given a model trained this way, inference is performed consistently by extracting pixel-wise embeddings and clustering, with the semantic label determined by the majority vote of its nearest neighbors from an annotated set. As a result, we present the SegSort, as a first attempt using deep learning for unsupervised semantic segmentation, achieving $76\%$ performance of its supervised counterpart. When supervision is available, SegSort shows consistent improvements over conventional approaches based on pixel-wise softmax training. Additionally, our approach produces more precise boundaries and consistent region predictions. The proposed SegSort further produces an interpretable result, as each choice of label can be easily understood from the retrieved nearest segments.

cs.CV↗
arXiv2019

Empirically Measuring Concentration: Fundamental Limits on Intrinsic Robustness

Many recent works have shown that adversarial examples that fool classifiers can be found by minimally perturbing a normal input. Recent theoretical results, starting with Gilmer et al. (2018b), show that if the inputs are drawn from a concentrated metric probability space, then adversarial examples with small perturbation are inevitable. A concentrated space has the property that any subset with $Ω(1)$ (e.g., 1/100) measure, according to the imposed distribution, has small distance to almost all (e.g., 99/100) of the points in the space. It is not clear, however, whether these theoretical results apply to actual distributions such as images. This paper presents a method for empirically measuring and bounding the concentration of a concrete dataset which is proven to converge to the actual concentration. We use it to empirically estimate the intrinsic robustness to $\ell_\infty$ and $\ell_2$ perturbations of several image classification benchmarks. Code for our experiments is available at https://github.com/xiaozhanguva/Measure-Concentration.

cs.LG↗
arXiv2019

Surface-charge-mediated Formation of H-TiO2@Ni(OH)2 Heterostructures for High-Performance Supercapcitors

Supercapacitors or ultracapacitors are promising for efficient energy storage applications, owing to their high power density, high charge-discharge rates, and long cycle life performance. To achieve this goal, a large specific surface area, an high electronic conductivity and a fast cation intercalation de-intercalation process are generally required in the design and preparation of materials for high-performance supercapacitors. Recently, core-shell heterostructures with multifunctionalities are regarded as one of promising materials for supercapacitors or ultracapacitors applications. In particular, one-dimensional (1D) core-shell heterostructures have sparked great scientific and technological interests due to their high versatility and applicability as the essential components in nanoscale electronics, catalysis, chemical sensing, and energy conversion storage devices. Various metal metal oxide, metal metal, metal oxide metal oxide and metal oxide conductive polymers so far have been investigated. Transition metal hydroxide oxide Co3O4, Co(OH)2, MnO2, Mn(OH)2, NiO, Ni(OH)2 and their compounds storing energy by surface faradaic (redox) reactions were generally integrated with conducting scaffold to build core-shell structure. Intensive studies show that an enlarged active surface area of transition metal hydroxide oxide enables a promoted surface redox reaction and enhanced electrochemical performance. Therefore, controllable synthesis of a hierarchically porous construction with high surface areas is critically important for energy storage.

cond-mat.mtrl-sci↗
arXiv2019

Utilization of Strong Charge Transfer Efficiency at 2H-1T Phase Boundary of MoS2 for Superb High-Performance Charge Storage

Transition metal dichalcogenides like MoS2 can exist many phases like the semiconducting 2H and the metallic 1T phases which have shown intriguing properties for energy and electrocatalytic applications. However, the 2H and 1T phases normally distribute coherently in a single-layered MoS2 sheet which is accompanied with ubiquitous hetero-phase boundaries. In this work, by using density functional theory and electrochemical measurement, we report strong charge transfer ability between 2H/1T phase boundary of MoS2 and graphene which accounts for a superb coexistence of gravimetric and volumetric capacitances of 272 F g-1 and 685 F cm-3. As a proof-of-concept application, a flexible solid-state asymmetric supercapacitor based on MoS2/graphene is fabricated, showing a remarkable energy and power densities (46.3 mWh cm-3 and 3.013 Wcm-3). Our work shows the promise of promoting the efficiency of charge flow and energy storage through engineering phase boundary and interface in phase-change materials.

physics.app-ph↗
arXiv2019

Low-cost LIDAR based Vehicle Pose Estimation and Tracking

Detecting surrounding vehicles by low-cost LIDAR has been drawing enormous attention. In low-cost LIDAR, vehicles present a multi-layer L-Shape. Based on our previous optimization/criteria-based L-Shape fitting algorithm, we here propose a data-driven and model-based method for robust vehicle segmentation and tracking. The new method uses T-linkage RANSAC to take a limited amount of noisy data and performs a robust segmentation for a moving car against noise. Compared with our previous method, T-Linkage RANSAC is more tolerant of observation uncertainties, i.e., the number of sides of the target being observed, and gets rid of the L-Shape assumption. In addition, a vehicle tracking system with Multi-Model Association (MMA) is built upon the segmentation result, which provides smooth trajectories of tracked objects. A manually labeled dataset from low-cost multi-layer LIDARs for validation will also be released with the paper. Experiments on the dataset show that the new approach outperforms previous ones based on multiple criteria. The new algorithm can also run in real-time.

cs.RO↗
arXiv2019

The structure of the blue whirl revealed

The blue whirl is a small, stable, spinning blue flame that evolved spontaneously in laboratory experiments while studying, violent, turbulent fire whirls. The blue whirl cleanly burns heavy, liquid hydrocarbon fuels with no soot production, presenting a new potential way for low-emission combustion. It is reproducible, appears for a range of different fuels and initial conditions, is quiet, appears laminar, and has characteristics which led to the idea that it results from vortex breakdown in whirling, reacting fluid. Since its discovery, considerable effort has been put into measurements, which have shown its temperature structure and sensitivity to the boundary layer near the surface. This has led to considerable speculation about the type of flames that comprise it. Simultaneously, there was a numerical effort to study its structure by performing simulations of vortex breakdown in gaseous reactive flows. The simulations described in this paper show that the stable blue whirl is composed of three different flame structures -- a diffusion flame and a premixed rich and lean flame -- all of which meet in a fourth structure, a triple flame which appears as a whirling blue ring. In addition, the blue whirl structure emerges as the result of vortex breakdown in a swirling reactive flow, as evidenced in the simulation by a bubble mode that is usually invisible in the experiments but at the center of the whirl. This paper also presents the tool used for the study and discusses how this might be used for future investigations.

physics.flu-dyn↗
arXiv2019

Energy-saving deployment algorithms of UAV swarm for sustainable wireless coverage

Recent years have witnessed increasingly more uses of Unmanned Aerial Vehicle (UAV) swarms for rapidly providing wireless coverage to ground users. Each UAV is constrained in its energy storage and wireless coverage, and it consumes most energy when flying to the top of the target area, leaving limited leftover energy for hovering at its deployed position and keeping wireless coverage. The literature largely overlooks this sustainability issue of deploying UAV swarm deployment, and we aim to maximize the minimum leftover energy storage among all the UAVs after their deployment. Our new energy-saving deployment problem captures that each UAV's wireless coverage is adjustable by its service altitude, and also takes the no-fly-zone (NFZ) constraint into account. Despite of this, we propose an optimal energy-saving deployment algorithm by jointly balancing heterogeneous UAVs' flying distances on the ground and final service altitudes in the sky. We show that a UAV with larger initial energy storage in the UAV swarm should be deployed further away from the UAV station. Moreover, when $n$ homogeneous UAVs are dispatched from different initial locations, we first prove that any two UAVs of the same initial energy storage will not fly across each other, and then design an approximation algorithm of complexity $n \log \frac{1}ε$ to arbitrarily approach the optimum with error $ε$. Finally, we consider that UAVs may have different initial energy storages, and we prove this problem is NP-hard. Despite of this, we successfully propose a heuristic algorithm to solve it by balancing the efficiency and computation complexity well.

cs.NI↗
arXiv2019

Enhanced higher harmonic generation from nodal topology

Among topological materials, nodal loop semimetals (NLSMs) are arguably the most topologically sophisticated, with their valence and conduction bands intersecting along arbitrarily intertwined nodes. But unlike the well-known topological band insulators with quantized edge conductivities, nodal loop materials possess topologically nontrivial Fermi surfaces, not bands. Hence an important question arises: Are there also directly measurable or even technologically useful physical properties characterizing nontrivial nodal loop topology? In this work, we provide an affirmative answer by showing, for the first time, that nodal linkages protect the higher harmonic generation (HHG) of electromagnetic signals. Specifically, nodal linkages enforce non-monotonicity in the intra-band semi-classical response of nodal materials, which will be robust against perturbations preserving the nodal topology. These nonlinearities distort incident radiation and produce higher frequency peaks in the teraHertz (THz) regime, as we quantitatively demonstrate for a few known nodal materials. Since THz sources are not yet ubiquitous, our new mechanism for HHG will greatly aid applications like material characterization and non-ionizing imaging of object interiors.

cond-mat.mes-hall↗