arXiv subjects

Yi Zhang

Publications and source records attributed to Yi Zhang.

At least 1,117 records · Page 62Linked to original sources

arXiv2019

Growth and Thermo-driven Crystalline Phase Transition of Metastable Monolayer 1T'-WSe2 Thin Film

Two-dimensional (2D) transition metal dichalcogenides MX2 (M = Mo, W, X = S, Se, Te) attracts enormous research interests in recent years. Its 2H phase possesses an indirect to direct bandgap transition in 2D limit, and thus shows great application potentials in optoelectronic devices [1]. The 1T' crystalline phase transition can drive the monolayer MX2 to be a 2D topological insulator. Here we realized the molecular beam epitaxial (MBE) growth of both the 1T' and 2H phase monolayer WSe2 on bilayer graphene (BLG) substrate. The crystalline structures of these two phases were characterized using scanning tunneling microscopy. The monolayer 1T'-WSe2 was found to be metastable, and can transform into 2H phase under post-annealing procedure. The phase transition temperature of 1T'-WSe2 grown on BLG is lower than that of 1T' phase grown on 2H-WSe2 layers. This thermo-driven crystalline phase transition makes the monolayer WSe2 to be an ideal platform for the controlling of topological phase transitions in 2D materials family.

cond-mat.mtrl-sci↗

arXiv2019

Extreme Tensoring for Low-Memory Preconditioning

State-of-the-art models are now trained with billions of parameters, reaching hardware limits in terms of memory consumption. This has created a recent demand for memory-efficient optimizers. To this end, we investigate the limits and performance tradeoffs of memory-efficient adaptively preconditioned gradient methods. We propose extreme tensoring for high-dimensional stochastic optimization, showing that an optimizer needs very little memory to benefit from adaptive preconditioning. Our technique applies to arbitrary models (not necessarily with tensor-shaped parameters), and is accompanied by regret and convergence guarantees, which shed light on the tradeoffs between preconditioner quality and expressivity. On a large-scale NLP model, we reduce the optimizer memory overhead by three orders of magnitude, without degrading performance.

cs.LG↗

arXiv2019

Multi-faceted machine learning of competing orders in disordered interacting systems

While the non-perturbative interaction effects in the fractional quantum Hall regime can be readily simulated through exact diagonalization, it has been challenging to establish a suitable diagnostic that can label different phases in the presence of competing interactions and disorder. Here we introduce a multi-faceted framework using a simple artificial neural network (ANN) to detect defining features of a fractional quantum Hall state, a charge density wave state and a localized state using the entanglement spectra and charge density as independent input. We consider the competing effects of a perturbing interaction ($l = 1$ pseudopotential $ΔV_1$), a disorder potential $W$, and the Coulomb interaction to the system at filling fraction $ν = 1/3$. Our phase diagram benchmarks well against previous estimates of the phase boundary using conventional measures along the $ΔV_1 = 0$ and $W = 0$ axes, the only regions where conventional approaches have been explored. Moreover, exploring the entire two-dimensional phase diagram for the first time, we establish the robustness of the fractional quantum Hall state and map out the charge density wave micro-emulsion phase wherein droplets of charge density wave region appear before the charge density wave is completely disordered. Hence we establish that the ANN can access and learn the defining traits of topological as well as broken symmetry phases using multi-faceted inputs of entanglement spectra and charge density.

cond-mat.str-el↗

arXiv2019

Optimizing seed inputs in fuzzing with machine learning

The success of a fuzzing campaign is heavily depending on the quality of seed inputs used for test generation. It is however challenging to compose a corpus of seed inputs that enable high code and behavior coverage of the target program, especially when the target program requires complex input formats such as PDF files. We present a machine learning based framework to improve the quality of seed inputs for fuzzing programs that take PDF files as input. Given an initial set of seed PDF files, our framework utilizes a set of neural networks to 1) discover the correlation between these PDF files and the execution in the target program, and 2) leverage such correlation to generate new seed files that more likely explore new paths in the target program. Our experiments on a set of widely used PDF viewers demonstrate that the improved seed inputs produced by our framework could significantly increase the code coverage of the target program and the likelihood of detecting program crashes.

cs.CR↗

arXiv2019

Cooperative Driving at Unsignalized Intersections Using Tree Search

In this paper, we propose a new cooperative driving strategy for connected and automated vehicles (CAVs) at unsignalized intersections. Based on the tree representation of the solution space for the passing order, we combine Monte Carlo tree search (MCTS) and some heuristic rules to find a nearly global-optimal passing order (leaf node) within a very short planning time. Testing results show that this new strategy can keep a good tradeoff between performance and computation flexibility.

cs.MA↗

arXiv2019

On Reinforcement Learning for Full-length Game of StarCraft

StarCraft II poses a grand challenge for reinforcement learning. The main difficulties of it include huge state and action space and a long-time horizon. In this paper, we investigate a hierarchical reinforcement learning approach for StarCraft II. The hierarchy involves two levels of abstraction. One is the macro-action automatically extracted from expert's trajectories, which reduces the action space in an order of magnitude yet remains effective. The other is a two-layer hierarchical architecture which is modular and easy to scale, enabling a curriculum transferring from simpler tasks to more complex tasks. The reinforcement training algorithm for this architecture is also investigated. On a 64x64 map and using restrictive units, we achieve a winning rate of more than 99\% against the difficulty level-1 built-in AI. Through the curriculum transfer learning algorithm and a mixture of combat model, we can achieve over 93\% winning rate of Protoss against the most difficult non-cheating built-in AI (level-7) of Terran, training within two days using a single machine with only 48 CPU cores and 8 K40 GPUs. It also shows strong generalization performance, when tested against never seen opponents including cheating levels built-in AI and all levels of Zerg and Protoss built-in AI. We hope this study could shed some light on the future research of large-scale reinforcement learning.

cs.LG↗

arXiv2019

Rational Solutions of First-Order Algebraic Ordinary Difference Equations

We propose an algebraic geometric approach for studying rational solutions of first-order algebraic ordinary difference equations. For an autonomous first-order algebraic ordinary difference equations, we give an upper bound for the degrees of its rational solutions, and thus derive a complete algorithm for computing corresponding rational solutions.

cs.SC↗

arXiv2019

Towards a sound massive cosmology

It is known that de Rham-Gabadadze-Tolley (dRGT) massive gravity does not permit a homogeneous and isotropic universe with flat or spherical spatial metrics. We demonstrate that a singular reference metric solves this problem in an economic and straightforward way. In the dRGT massive gravity with a singular reference metric, there are sound homogeneous and isotropic cosmological solutions. We investigate cosmologies with the static and dynamical singular reference metrics, respectively. The term like dark energy appears naturally and the universe accelerates itself in some late time evolution. The term simulating dark matter also naturally emerges. We make a preliminary constraint on the parameters in the dRGT massive gravity in frame of the present cosmological model by using the data of supernovae, cosmic microwave back ground radiations, and baryonic acoustic oscillations.

gr-qc↗

arXiv2019

Three-dimensional Optical Coherence Tomography Image Denoising through Multi-input Fully-Convolutional Networks

In recent years, there has been a growing interest in applying convolutional neural networks (CNNs) to low-level vision tasks such as denoising and super-resolution. Due to the coherent nature of the image formation process, optical coherence tomography (OCT) images are inevitably affected by noise. This paper proposes a new method named the multi-input fully-convolutional networks (MIFCN) for denoising of OCT images. In contrast to recently proposed natural image denoising CNNs, the proposed architecture allows the exploitation of high degrees of correlation and complementary information among neighboring OCT images through pixel by pixel fusion of multiple FCNs. The parameters of the proposed multi-input architecture are learned by considering the consistency between the overall output and the contribution of each input image. The proposed MIFCN method is compared with the state-of-the-art denoising methods adopted on OCT images of normal and age-related macular degeneration eyes in a quantitative and qualitative manner.

cs.CV↗

arXiv2019

Vertex degree sums for matchings in 3-uniform hypergraphs

Let $n, s$ be positive integers such that $n$ is sufficiently large and $s\le n/3$. Suppose $H$ is a 3-uniform hypergraph of order $n$. If $H$ contains no isolated vertex and $deg(u)+ deg(v) > 2(s-1)(n-1)$ for any two vertices $u$ and $v$ that are contained in some edge of $H$, then $H$ contains a matching of size $s$. This degree sum condition is best possible and confirms a conjecture of the authors [Electron. J. Combin. 25 (3), 2018], who proved the case when $s= n/3$.

math.CO↗

arXiv2018

TechKG: A Large-Scale Chinese Technology-Oriented Knowledge Graph

Knowledge graph is a kind of valuable knowledge base which would benefit lots of AI-related applications. Up to now, lots of large-scale knowledge graphs have been built. However, most of them are non-Chinese and designed for general purpose. In this work, we introduce TechKG, a large scale Chinese knowledge graph that is technology-oriented. It is built automatically from massive technical papers that are published in Chinese academic journals of different research domains. Some carefully designed heuristic rules are used to extract high quality entities and relations. Totally, it comprises of over 260 million triplets that are built upon more than 52 million entities which come from 38 research domains. Our preliminary ex-periments indicate that TechKG has high adaptability and can be used as a dataset for many diverse AI-related applications. We released TechKG at: http://www.techkg.cn.

cs.AI↗

arXiv2018

Probing transport in quantum many-fermion simulations via quantum loop topography

Quantum many-fermion systems give rise to diverse states of matter that often reveal themselves in distinctive transport properties. While some of these states can be captured by microscopic models accessible to numerical exact quantum Monte Carlo simulations, it nevertheless remains challenging to numerically access their transport properties. Here we demonstrate that quantum loop topography (QLT) can be used to directly probe transport by machine learning current-current correlations in imaginary time. We showcase this approach by studying the emergence of superconducting fluctuations in the negative-U Hubbard model and a spin-fermion model for a metallic quantum critical point. For both sign-free models, we find that the QLT approach detects a change in transport in very good agreement with their established phase diagrams. These proof-of-principle calculations combined with the numerical efficiency of the QLT approach point a way to identify hitherto elusive transport phenomena such as non-Fermi liquids using machine learning algorithms.

cond-mat.str-el↗

arXiv2018

Finite horizon risk-sensitive continuous-time Markov decision processes with unbounded transition and cost rates

We consider a risk-sensitive continuous-time Markov decision process over a finite time duration. Under the conditions that can be satisfied by unbounded transition and cost rates, we show the existence of an optimal policy, and the existence and uniqueness of the solution to the optimality equation out of a class of possibly unbounded functions, to which the Feynman-Kac formula was also justified to hold.

math.OC↗

arXiv2018

Stronger generalization bounds for deep nets via a compression approach

Deep nets generalize well despite having more parameters than the number of training samples. Recent works try to give an explanation using PAC-Bayes and Margin-based analyses, but do not as yet result in sample complexity bounds better than naive parameter counting. The current paper shows generalization bounds that're orders of magnitude better in practice. These rely upon new succinct reparametrizations of the trained net --- a compression that is explicit and efficient. These yield generalization bounds via a simple compression-based framework introduced here. Our results also provide some theoretical justification for widespread empirical success in compressing deep nets. Analysis of correctness of our compression relies upon some newly identified \textquotedblleft noise stability\textquotedblright properties of trained deep nets, which are also experimentally verified. The study of these properties and resulting generalization bounds are also extended to convolutional nets, which had eluded earlier attempts on proving generalization.

cs.LG↗

arXiv2018

Strong convergence of a fully discrete finite element method for a class of semilinear stochastic partial differential equations with multiplicative noise

This paper develops and analyzes a fully discrete finite element method for a class of semilinear stochastic partial differential equations (SPDEs) with multiplicative noise. The nonlinearity in the diffusion term of the SPDEs is assumed to be globally Lipschitz and the nonlinearity in the drift term is only assumed to satisfy a one-side Lipschitz condition. The semilinear SPDEs considered in this paper is a direct generalization of the SODEs considered in [13]. There are several difficulties which need to be overcome for this generalization. First, obviously the spatial discretization, which does not appear in the SODE case, adds an extra layer of difficulty. It turns out a special discretization must be designed to guarantee certain properties for the numerical scheme and its stiffness matrix. In this paper we use a finite element interpolation technique to discretize the nonlinear drift term. Second, in order to prove the strong convergence of the proposed fully discrete finite element method, stability estimates for higher order moments of the $H^1$-seminorm of the numerical solution must be established, which are difficult and delicate. A judicious combination of the properties of the drift and diffusion terms and a nontrivial technique borrowed from [16] is used in this paper to achieve the goal. Finally, stability estimates for the second and higher order moments of the $L^2$-norm of the numerical solution is also difficult to obtain due to the fact that the mass matrix may not be diagonally dominant. This is done by utilizing the interpolation theory and the higher moment estimates for the $H^1$-seminorm of the numerical solution. After overcoming these difficulties, it is proved that the proposed fully discrete finite element method is convergent in strong norms with nearly optimal rates of convergence.

math.NA↗

arXiv2018

Deep Unfolded Robust PCA with Application to Clutter Suppression in Ultrasound

Contrast enhanced ultrasound is a radiation-free imaging modality which uses encapsulated gas microbubbles for improved visualization of the vascular bed deep within the tissue. It has recently been used to enable imaging with unprecedented subwavelength spatial resolution by relying on super-resolution techniques. A typical preprocessing step in super-resolution ultrasound is to separate the microbubble signal from the cluttering tissue signal. This step has a crucial impact on the final image quality. Here, we propose a new approach to clutter removal based on robust principle component analysis (PCA) and deep learning. We begin by modeling the acquired contrast enhanced ultrasound signal as a combination of a low rank and sparse components. This model is used in robust PCA and was previously suggested in the context of ultrasound Doppler processing and dynamic magnetic resonance imaging. We then illustrate that an iterative algorithm based on this model exhibits improved separation of microbubble signal from the tissue signal over commonly practiced methods. Next, we apply the concept of deep unfolding to suggest a deep network architecture tailored to our clutter filtering problem which exhibits improved convergence speed and accuracy with respect to its iterative counterpart. We compare the performance of the suggested deep network on both simulations and in-vivo rat brain scans, with a commonly practiced deep-network architecture and the fast iterative shrinkage algorithm, and show that our architecture exhibits better image quality and contrast.

cs.LG↗

arXiv2018

Towards Easier and Faster Sequence Labeling for Natural Language Processing: A Search-based Probabilistic Online Learning Framework (SAPO)

There are two major approaches for sequence labeling. One is the probabilistic gradient-based methods such as conditional random fields (CRF) and neural networks (e.g., RNN), which have high accuracy but drawbacks: slow training, and no support of search-based optimization (which is important in many cases). The other is the search-based learning methods such as structured perceptron and margin infused relaxed algorithm (MIRA), which have fast training but also drawbacks: low accuracy, no probabilistic information, and non-convergence in real-world tasks. We propose a novel and "easy" solution, a search-based probabilistic online learning method, to address most of those issues. The method is "easy", because the optimization algorithm at the training stage is as simple as the decoding algorithm at the test stage. This method searches the output candidates, derives probabilities, and conducts efficient online learning. We show that this method with fast training and theoretical guarantee of convergence, which is easy to implement, can support search-based optimization and obtain top accuracy. Experiments on well-known tasks show that our method has better accuracy than CRF and BiLSTM\footnote{The SAPO code is released at \url{https://github.com/lancopku/SAPO}.}.

cs.LG↗

arXiv2018

Observation of Topologically Protected States at Crystalline Phase Boundaries in Single-layer WSe2

Transition metal dichalcogenide (TMD) materials are unique in the wide variety of structural and electronic phases they exhibit in the two-dimensional (2D) single-layer limit. Here we show how such polymorphic flexibility can be used to achieve topological states at highly ordered phase boundaries in a new quantum spin Hall insulator (QSHI), 1T'-WSe2. We observe helical states at the crystallographically-aligned interface between quantum a spin Hall insulating domain of 1T'-WSe2 and a semiconducting domain of 1H-WSe2 in contiguous single layers grown using molecular beam epitaxy (MBE). The QSHI nature of single-layer 1T'-WSe2 was verified using ARPES to determine band inversion around a 120 meV energy gap, as well as STM spectroscopy to directly image helical edge-state formation. Using this new edge-state geometry we are able to directly confirm the predicted penetration depth of a helical interface state into the 2D bulk of a QSHI for a well-specified crystallographic direction. The clean, well-ordered topological/trivial interfaces observed here create new opportunities for testing predictions of the microscopic behavior of topologically protected boundary states without the complication of structural disorder.

cond-mat.mtrl-sci↗