arXiv subjects

Lijun Zhang

Publications and source records attributed to Lijun Zhang.

At least 217 records · Page 12Linked to original sources

arXiv2020

New Polymorphs of Two-Dimensional Indium Selenide with Enhanced Electronic Properties

The two-dimensional (2D) semiconductor indium selenide (InSe) has attracted significant interest due its unique electronic band structure, high electron mobility and wide tunability of its band gap energy achieved by varying the layer thickness. All these features make 2D InSe a potential candidate for advanced electronic and optoelectronic applications. Here, we report on the discovery of new polymorphs of InSe with enhanced electronic properties. Using a global structure search that combines artificial swarm intelligence with first-principles energetic calculations, we identify polymorphs that consist of a centrosymmetric monolayer belonging to the point group D$_{3d}$, distinct from the well-known polymorphs based on the D$_{3h}$ monolayers that lack inversion symmetry. The new polymorphs are thermodynamically and kinetically stable, and exhibit a wider optical spectral response and larger electron mobilities compared to the known polymorphs. We discuss opportunities to synthesize these newly discovered polymorphs and viable routes to identify them by X-ray diffraction, Raman spectroscopy and second harmonic generation experiments.

cond-mat.mtrl-sci↗

arXiv2020

Nearly Optimal Regret for Stochastic Linear Bandits with Heavy-Tailed Payoffs

In this paper, we study the problem of stochastic linear bandits with finite action sets. Most of existing work assume the payoffs are bounded or sub-Gaussian, which may be violated in some scenarios such as financial markets. To settle this issue, we analyze the linear bandits with heavy-tailed payoffs, where the payoffs admit finite $1+ε$ moments for some $ε\in(0,1]$. Through median of means and dynamic truncation, we propose two novel algorithms which enjoy a sublinear regret bound of $\widetilde{O}(d^{\frac{1}{2}}T^{\frac{1}{1+ε}})$, where $d$ is the dimension of contextual information and $T$ is the time horizon. Meanwhile, we provide an $Ω(d^{\fracε{1+ε}}T^{\frac{1}{1+ε}})$ lower bound, which implies our upper bound matches the lower bound up to polylogarithmic factors in the order of $d$ and $T$ when $ε=1$. Finally, we conduct numerical experiments to demonstrate the effectiveness of our algorithms and the empirical results strongly support our theoretical guarantees.

cs.LG↗

arXiv2020

Bandit Convex Optimization in Non-stationary Environments

Bandit Convex Optimization (BCO) is a fundamental framework for modeling sequential decision-making with partial information, where the only feedback available to the player is the one-point or two-point function values. In this paper, we investigate BCO in non-stationary environments and choose the \emph{dynamic regret} as the performance measure, which is defined as the difference between the cumulative loss incurred by the algorithm and that of any feasible comparator sequence. Let $T$ be the time horizon and $P_T$ be the path-length of the comparator sequence that reflects the non-stationarity of environments. We propose a novel algorithm that achieves $O(T^{3/4}(1+P_T)^{1/2})$ and $O(T^{1/2}(1+P_T)^{1/2})$ dynamic regret respectively for the one-point and two-point feedback models. The latter result is optimal, matching the $Ω(T^{1/2}(1+P_T)^{1/2})$ lower bound established in this paper. Notably, our algorithm is more adaptive to non-stationary environments since it does not require prior knowledge of the path-length $P_T$ ahead of time, which is generally unknown.

cs.LG↗

arXiv2020

Adaptive and Efficient Algorithms for Tracking the Best Expert

In this paper, we consider the problem of prediction with expert advice in dynamic environments. We choose tracking regret as the performance metric and develop two adaptive and efficient algorithms with data-dependent tracking regret bounds. The first algorithm achieves a second-order tracking regret bound, which improves existing first-order bounds. The second algorithm enjoys a path-length bound, which is generally not comparable to the second-order bound but offers advantages in slowly moving environments. Both algorithms are developed under the online mirror descent framework and draw inspiration from existing algorithms that attain data-dependent bounds of static regret. The key idea is to use a clipped simplex in the updating step of online mirror descent. Finally, we extend our algorithms and analysis to online matrix prediction and provide the first data-dependent tracking regret bound for this problem.

cs.LG↗

arXiv2020

Minimizing Dynamic Regret and Adaptive Regret Simultaneously

Regret minimization is treated as the golden rule in the traditional study of online learning. However, regret minimization algorithms tend to converge to the static optimum, thus being suboptimal for changing environments. To address this limitation, new performance measures, including dynamic regret and adaptive regret have been proposed to guide the design of online algorithms. The former one aims to minimize the global regret with respect to a sequence of changing comparators, and the latter one attempts to minimize every local regret with respect to a fixed comparator. Existing algorithms for dynamic regret and adaptive regret are developed independently, and only target one performance measure. In this paper, we bridge this gap by proposing novel online algorithms that are able to minimize the dynamic regret and adaptive regret simultaneously. In fact, our theoretical guarantee is even stronger in the sense that one algorithm is able to minimize the dynamic regret over any interval.

cs.LG↗

arXiv2019

Stochastic Primal-Dual Algorithms with Faster Convergence than $O(1/\sqrt{T})$ for Problems without Bilinear Structure

Previous studies on stochastic primal-dual algorithms for solving min-max problems with faster convergence heavily rely on the bilinear structure of the problem, which restricts their applicability to a narrowed range of problems. The main contribution of this paper is the design and analysis of new stochastic primal-dual algorithms that use a mixture of stochastic gradient updates and a logarithmic number of deterministic dual updates for solving a family of convex-concave problems with no bilinear structure assumed. Faster convergence rates than $O(1/\sqrt{T})$ with $T$ being the number of stochastic gradient updates are established under some mild conditions of involved functions on the primal and the dual variable. For example, for a family of problems that enjoy a weak strong convexity in terms of the primal variable and has a strongly concave function of the dual variable, the convergence rate of the proposed algorithm is $O(1/T)$. We also investigate the effectiveness of the proposed algorithms for learning robust models and empirical AUC maximization.

cs.LG↗

arXiv2019

Interdiffusion between gadolinia doped ceria and yttria stabilized zirconia in solid oxide fuel cells: experimental investigation and kinetic modeling

Interdiffusion between the yttria stabilized zirconia (YSZ) electrolyte and the gadolinia doped ceria (CGO) barrier layer is one of the major causes to the degradation of solid oxide fuel cells (SOFCs). We present in this work experimental investigations on CGO-YSZ bi-layer electrolyte sintered at 1250 C or 1315 C and element transport as a function of sintering temperature and dwelling time. In order to quantitatively simulate the experimental observations, the CALPHAD-type thermodynamic assessment of the CGO-YSZ system is performed by simplifying the system to a CeO2-ZrO2 quasi-binary system, and the kinetic descriptions (atomic mobilities) are constructed based on critical review of literature data. The CGO-YSZ interdiffusion is then modeled with the DICTRA software and the simulation results are compared with the experimental data under different sintering or long-term operating conditions. The corresponding ohmic resistance of the bi-layer electrolyte is predicted based on the simulated concentration profile. The results implies that the interdiffusion across the CGO-YSZ interface happens mainly during sintering at high temperature, while during long-term operation at relatively lower temperature the impact of interdiffusion on cell degradation is negligible.

cond-mat.mtrl-sci↗

arXiv2019

Multi-attention Networks for Temporal Localization of Video-level Labels

Temporal localization remains an important challenge in video understanding. In this work, we present our solution to the 3rd YouTube-8M Video Understanding Challenge organized by Google Research. Participants were required to build a segment-level classifier using a large-scale training data set with noisy video-level labels and a relatively small-scale validation data set with accurate segment-level labels. We formulated the problem as a multiple instance multi-label learning and developed an attention-based mechanism to selectively emphasize the important frames by attention weights. The model performance is further improved by constructing multiple sets of attention networks. We further fine-tuned the model using the segment-level data set. Our final model consists of an ensemble of attention/multi-attention networks, deep bag of frames models, recurrent neural networks and convolutional neural networks. It ranked 13th on the private leader board and stands out for its efficient usage of resources.

cs.CV↗

arXiv2019

Adaptive Regret of Convex and Smooth Functions

We investigate online convex optimization in changing environments, and choose the adaptive regret as the performance measure. The goal is to achieve a small regret over every interval so that the comparator is allowed to change over time. Different from previous works that only utilize the convexity condition, this paper further exploits smoothness to improve the adaptive regret. To this end, we develop novel adaptive algorithms for convex and smooth functions, and establish problem-dependent regret bounds over any interval. Our regret bounds are comparable to existing results in the worst case, and become much tighter when the comparator has a small loss.

cs.LG↗

arXiv2019

Multi-Objective Generalized Linear Bandits

In this paper, we study the multi-objective bandits (MOB) problem, where a learner repeatedly selects one arm to play and then receives a reward vector consisting of multiple objectives. MOB has found many real-world applications as varied as online recommendation and network routing. On the other hand, these applications typically contain contextual information that can guide the learning process which, however, is ignored by most of existing work. To utilize this information, we associate each arm with a context vector and assume the reward follows the generalized linear model (GLM). We adopt the notion of Pareto regret to evaluate the learner's performance and develop a novel algorithm for minimizing it. The essential idea is to apply a variant of the online Newton step to estimate model parameters, based on which we utilize the upper confidence bound (UCB) policy to construct an approximation of the Pareto front, and then uniformly at random choose one arm from the approximate Pareto front. Theoretical analysis shows that the proposed algorithm achieves an $\tilde O(d\sqrt{T})$ Pareto regret, where $T$ is the time horizon and $d$ is the dimension of contexts, which matches the optimal result for single objective contextual bandits problem. Numerical experiments demonstrate the effectiveness of our method.

cs.LG↗

arXiv2019

Improving the Robustness of Deep Neural Networks via Adversarial Training with Triplet Loss

Recent studies have highlighted that deep neural networks (DNNs) are vulnerable to adversarial examples. In this paper, we improve the robustness of DNNs by utilizing techniques of Distance Metric Learning. Specifically, we incorporate Triplet Loss, one of the most popular Distance Metric Learning methods, into the framework of adversarial training. Our proposed algorithm, Adversarial Training with Triplet Loss (AT$^2$L), substitutes the adversarial example against the current model for the anchor of triplet loss to effectively smooth the classification boundary. Furthermore, we propose an ensemble version of AT$^2$L, which aggregates different attack methods and model structures for better defense effects. Our empirical studies verify that the proposed approach can significantly improve the robustness of DNNs without sacrificing accuracy. Finally, we demonstrate that our specially designed triplet loss can also be used as a regularization term to enhance other defense methods.

cs.LG↗

arXiv2019

Adaptivity and Optimality: A Universal Algorithm for Online Convex Optimization

In this paper, we study adaptive online convex optimization, and aim to design a universal algorithm that achieves optimal regret bounds for multiple common types of loss functions. Existing universal methods are limited in the sense that they are optimal for only a subclass of loss functions. To address this limitation, we propose a novel online method, namely Maler, which enjoys the optimal $O(\sqrt{T})$, $O(d\log T)$ and $O(\log T)$ regret bounds for general convex, exponentially concave, and strongly convex functions respectively. The essential idea is to run multiple types of learning algorithms with different learning rates in parallel, and utilize a meta algorithm to track the best one on the fly. Empirical results demonstrate the effectiveness of our method.

cs.LG↗

arXiv2019

SAdam: A Variant of Adam for Strongly Convex Functions

The Adam algorithm has become extremely popular for large-scale machine learning. Under convexity condition, it has been proved to enjoy a data-dependant $O(\sqrt{T})$ regret bound where $T$ is the time horizon. However, whether strong convexity can be utilized to further improve the performance remains an open problem. In this paper, we give an affirmative answer by developing a variant of Adam (referred to as SAdam) which achieves a data-dependant $O(\log T)$ regret bound for strongly convex functions. The essential idea is to maintain a faster decaying yet under controlled step size for exploiting strong convexity. In addition, under a special configuration of hyperparameters, our SAdam reduces to SC-RMSprop, a recently proposed variant of RMSprop for strongly convex functions, for which we provide the first data-dependent logarithmic regret bound. Empirical results on optimizing strongly convex functions and training deep networks demonstrate the effectiveness of our method.

cs.LG↗

arXiv2019

Computational Design of Mixed-valence Tin Sulfides as Solar Absorbers

Binary tin sulfides are appealing because of their simple stoichiometry and semiconducting properties and are potentially cost-effective optoelectronic materials. The multivalency of Sn allows yet more intermediate compositions, Sn$_x$S$_y$, whose structures and properties are of interest.Sn$_2$S$_3$ is already under consideration as a mixed-valence semiconductor. Other intermediate compositions have remained elusive. Here we report a comprehensive study of phase stability of the Sn$_x$S$_y$ series compounds, utilizing swarm-intelligence crystal structure search method combined with first-principles energetic calculations. We find that the stability of mixed-valence Sn$_x$S$_y$ compounds with respect to decomposition into pure-valence SnS and SnS$_2$ is in general weaker than the Sn$_x$O$_y$ counterparts, likely due to differences in chemical bonding. Besides identifying the experimentally discovered stable phases of Sn$_2$S$_3$, our calculations indicate that the Sn$_3$S$_4$ phase is another mixed-valence composition which shows marginal stability with respect to decomposition into SnS and SnS$_2$. Other studied compositions may be metastable under ambient conditions, with slightly positive formation enthalpies. We find two structures of Sn$_3$S$_4$ having comparably low energies, both of which feature one-dimensional chain-like fragments obtained by breaking up the edge-connected octahedral layers of SnS$_2$. Both structures indicate lattice phonon stability and one shows quasi-direct band gap with a calculated value of 1.43 eV, ideal for solar absorbers. A further analysis of the composition-structure-property relationship supports the notion that lowdimensional Sn-S motifs and van der Waals interaction may lead to diverse structure types and chemical compositions, having functional properties that are yet to be identified in the Sn$_x$S$_y$ series with mixed valency.

cond-mat.mtrl-sci↗

arXiv2019

Dielectric Behavior as a Screen in Rational Searches for Electronic Materials: Metal Pnictide Sulfosalts

Dielectric screening plays an important role in reducing the strength of carrier scattering and trapping by point defects for many semiconductors such as the halide perovskite solar materials. However, it was rarely considered as a screen to find new electronic semiconductors. We performed a material search study using the dielectric properties as a screen to identify potential electronic materials in the class of metal-pnictide ternary sulfosalts, containing Bi or Sb. These salts are basically ionic due to the electronegativity difference between the S and both the metal and pnictogen elements. However, we do find significant cross-gap hybridization between the S p-derived valence bands and pnictogen p-derived conduction bands in many of the materials. This leads to enhanced Born effective charges, and in several cases, highly enhanced dielectric constants. We find a chemical rule for high dielectric constants in terms of the bond connectivity of the pnictogen-chalcogen part of the crystal structure. We additionally find a series of compounds with low effective mass, high dielectric constant and other properties that suggest good performance as electronic materials and also several potential thermoelectric compounds. Experimental optical data and solar conversion efficiency are reported for Sn-Sb-S samples, and results in accord with predicted good performance are found. The results illustrate the utility of dielectric properties as a screen for identifying complex semiconductors.

cond-mat.mtrl-sci↗

arXiv2019

Switchable out-of-plane polarization in two-dimensional LiAlTe$_2$

Covalent-polar semiconductors that show intrinsic two-dimensional (2D) vertical polarization present new device opportunities. These materials differ from ordinary ferroelectrics in that they are able to maintain polarization normal to a surface even with an unscreened depolarization field. Identifying phases that exhibit intrinsic 2D vertical polarization is an ongoing challenge. Here we report via computational material design the discovery of a new promising phase, specifically 2D LiAlTe$_2$. The design idea is developed from the physical understanding of three-dimensional hyperferroelectric covalent polar semiconductors. We used the structure determination method combining swarm intelligence algorithm and first-principles calculations to identify energetically stable structures. In addition to the expected layered version of bulk LiAlTe$_2$, $β$-LiAlTe$_2$, we find a novel 2D structure, $γ$-LiAlTe$_2$. In this phase, the vertical dipole can be switched between 0.07 and -0.11 e$\cdot$angstrom. This switching is triggered by the movement of Li atom between two local energy minima. The associated asymmetric double-well energy profile can be continuously tuned by the applied electric field as well as strain. There is, therefore, a reversible transition between two polar states. This discovered off-plane switchability provides an opportunity for the 2D $γ$-LiAlTe$_2$ based interfacial phase change memory device for example by growing $γ$-LiAlTe$_2$/GeTe heterostructures.

cond-mat.mtrl-sci↗

arXiv2019

Empirical-likelihood-based criteria for model selection on marginal analysis of longitudinal data with dropout missingness

Longitudinal data are common in clinical trials and observational studies, where missing outcomes due to dropouts are always encountered. Under such context with the assumption of missing at random, the weighted generalized estimating equations (WGEE) approach is widely adopted for marginal analysis. Model selection on marginal mean regression is a crucial aspect of data analysis, and identifying an appropriate correlation structure for model fitting may also be of interest and importance. However, the existing information criteria for model selection in WGEE have limitations, such as separate criteria for the selection of marginal mean and correlation structures, unsatisfactory selection performance in small-sample set-ups and so on. In particular, there are few studies to develop joint information criteria for selection of both marginal mean and correlation structures. In this work, by embedding empirical likelihood into the WGEE framework, we propose two innovative information criteria named a joint empirical Akaike information criterion (JEAIC) and a joint empirical Bayesian information criterion (JEBIC), which can simultaneously select the variables for marginal mean regression and also correlation structure. Through extensive simulation studies, these empirical-likelihood-based criteria exhibit robustness, flexibility, and outperformance compared to the other criteria including the weighted quasi-likelihood under the independence model criterion, the missing longitudinal information criterion and the joint longitudinal information criterion. In addition, we provide a theoretical justification of our proposed criteria, and present two real data examples in practice for further illustration.

stat.ME↗

arXiv2019

Stochastic Approximation of Smooth and Strongly Convex Functions: Beyond the $O(1/T)$ Convergence Rate

Stochastic approximation (SA) is a classical approach for stochastic convex optimization. Previous studies have demonstrated that the convergence rate of SA can be improved by introducing either smoothness or strong convexity condition. In this paper, we make use of smoothness and strong convexity simultaneously to boost the convergence rate. Let $λ$ be the modulus of strong convexity, $κ$ be the condition number, $F_*$ be the minimal risk, and $α>1$ be some small constant. First, we demonstrate that, in expectation, an $O(1/[λT^α] + κF_*/T)$ risk bound is attainable when $T = Ω(κ^α)$. Thus, when $F_*$ is small, the convergence rate could be faster than $O(1/[λT])$ and approaches $O(1/[λT^α])$ in the ideal case. Second, to further benefit from small risk, we show that, in expectation, an $O(1/2^{T/κ}+F_*)$ risk bound is achievable. Thus, the excess risk reduces exponentially until reaching $O(F_*)$, and if $F_*=0$, we obtain a global linear convergence. Finally, we emphasize that our proof is constructive and each risk bound is equipped with an efficient stochastic algorithm attaining that bound.

cs.LG↗