arXiv subjects

Yi Zhou

Publications and source records attributed to Yi Zhou.

At least 469 records · Page 26Linked to original sources

arXiv2020

Discovery of two-dimensional anisotropic superconductivity at KTaO$_3$ (111) interfaces

The unique electronic structure found at interfaces between materials can allow unconventional quantum states to emerge. Here we observe superconductivity in electron gases formed at interfaces between (111) oriented KTaO$_3$ and insulating overlayers of either EuO or LaAlO$_3$. The superconducting transition temperature, approaching 2.2 K, is about one order of magnitude higher than that of the LaAlO$_3$/SrTiO$_3$ system. Strikingly, similar electron gases at (001) KTaO$_3$ interfaces remain normal down to 25 mK. The critical field and current-voltage measurements indicate that the superconductivity is two dimensional. Higher mobility EuO/KTaO$_3$ (111) samples show a large in-plane anisotropy in transport properties at low temperatures prior to onset of superconductivity, suggesting the emergence of a stripe like phase where the superconductivity is nearly homogeneous in one direction, but strongly modulated in the other.

cond-mat.supr-con↗

arXiv2020

An Investigation into the Stochasticity of Batch Whitening

Batch Normalization (BN) is extensively employed in various network architectures by performing standardization within mini-batches. A full understanding of the process has been a central target in the deep learning communities. Unlike existing works, which usually only analyze the standardization operation, this paper investigates the more general Batch Whitening (BW). Our work originates from the observation that while various whitening transformations equivalently improve the conditioning, they show significantly different behaviors in discriminative scenarios and training Generative Adversarial Networks (GANs). We attribute this phenomenon to the stochasticity that BW introduces. We quantitatively investigate the stochasticity of different whitening transformations and show that it correlates well with the optimization behaviors during training. We also investigate how stochasticity relates to the estimation of population statistics during inference. Based on our analysis, we provide a framework for designing and comparing BW algorithms in different scenarios. Our proposed BW algorithm improves the residual networks by a significant margin on ImageNet classification. Besides, we show that the stochasticity of BW can improve the GAN's performance with, however, the sacrifice of the training stability.

cs.CV↗

arXiv2020

GFTE: Graph-based Financial Table Extraction

Tabular data is a crucial form of information expression, which can organize data in a standard structure for easy information retrieval and comparison. However, in financial industry and many other fields tables are often disclosed in unstructured digital files, e.g. Portable Document Format (PDF) and images, which are difficult to be extracted directly. In this paper, to facilitate deep learning based table extraction from unstructured digital files, we publish a standard Chinese dataset named FinTab, which contains more than 1,600 financial tables of diverse kinds and their corresponding structure representation in JSON. In addition, we propose a novel graph-based convolutional neural network model named GFTE as a baseline for future comparison. GFTE integrates image feature, position feature and textual feature together for precise edge prediction and reaches overall good results.

cs.CV↗

arXiv2020

Timing Performance of a Micro-Channel-Plate Photomultiplier Tube

The spatial dependence of the timing performance of the R3809U-50 Micro-Channel-Plate PMT (MCP-PMT) by Hamamatsu was studied in high energy muon beams. Particle position information is provided by a GEM tracker telescope, while timing is measured relative to a second MCP-PMT, identical in construction. In the inner part of the circular active area (radius r$<$5.5\,mm) the time resolution of the two MCP-PMTs combined is better than 10~ps. The signal amplitude decreases in the outer region due to less light reaching the photocathode, resulting in a worse time resolution. The observed radial dependence is in quantitative agreement with a dedicated simulation. With this characterization, the suitability of MCP-PMTs as $\text{t}_\text{0}$ reference detectors has been validated.

physics.ins-det↗

arXiv2020

Classical and quantum order in hyperkagome antiferromagnets

Motivated by recent experiments and density functional theory calculations on choloalite PbCuTe$_2$O$_6$, which possesses a Cu-based three-dimensional hyperkagome lattice, we propose and study a $J_1$-$J_2$-$J_3$ antiferromagnetic Heisenberg model on a hyperkagome lattice. In the classical limit, possible ground states are analyzed by two triangle rules, i.e., the "hyperkagome triangle rule" and the "isolated triangle rule," and classical Monte Carlo simulations are exploited to identify possible classical magnetic ordering and explore the phase diagram. In the quantum regime, Schwinger boson theory is applied to study possible quantum spin liquid states and long-range magnetically ordered states on an equal footing. These quantum states with bosonic partons are classified and analyzed by using projective symmetry groups (PSGs). It is found that there are only four types of algebraic PSGs allowed by the space group $P4_{1}32$ on a hyperkagome lattice. Moreover, there are only two types of PSGs that are compatible with the $J_1$-$J_2$-$J_3$ Heisenberg model. These two types of $Z_2$ bosonic states are distinguished by the gauge-invariant flux on the elementary ten-site loops on the hyperkagome network, called zero-flux state and $π$-flux state respectively. Both the zero-flux state and the $π$-flux state are able to give rise to quantum spin liquid states as well as magnetically ordered states, and the zero-flux states and the $π$-flux states can be distinguished by the lower and upper edges of the spectral function $S(\bm{q},ω)$, which can be measured by inelastic neutron scattering experiments.

cond-mat.str-el↗

arXiv2020

TiFL: A Tier-based Federated Learning System

Federated Learning (FL) enables learning a shared model across many clients without violating the privacy requirements. One of the key attributes in FL is the heterogeneity that exists in both resource and data due to the differences in computation and communication capacity, as well as the quantity and content of data among different clients. We conduct a case study to show that heterogeneity in resource and data has a significant impact on training time and model accuracy in conventional FL systems. To this end, we propose TiFL, a Tier-based Federated Learning System, which divides clients into tiers based on their training performance and selects clients from the same tier in each training round to mitigate the straggler problem caused by heterogeneity in resource and data quantity. To further tame the heterogeneity caused by non-IID (Independent and Identical Distribution) data and resources, TiFL employs an adaptive tier selection approach to update the tiering on-the-fly based on the observed training performance and accuracy overtime. We prototype TiFL in a FL testbed following Google's FL architecture and evaluate it using popular benchmarks and the state-of-the-art FL benchmark LEAF. Experimental evaluation shows that TiFL outperforms the conventional FL in various heterogeneous conditions. With the proposed adaptive tier selection policy, we demonstrate that TiFL achieves much faster training performance while keeping the same (and in some cases - better) test accuracy across the board.

cs.LG↗

arXiv2020

Measurement of the neutron beam profile of the Back-n white neutron facility at CSNS with a Micromegas detector

The Back-n white neutron beam line, which uses back-streaming white neutrons from the spallation target of the China Spallation Neutron Source, is used for nuclear data measurements. A Micromegas-based neutron detector with two variants was specially developed to measure the beam spot distribution for this beam line. In this article, the design, fabrication, and characterization of the detector are described. The results of the detector performance tests are presented, which include the relative electron transparency, the gain and the gain uniformity, and the neutron beam profile reconstruction capability. The result of the first measurement of the Back-n neutron beam spot distribution is also presented.

physics.ins-det↗

arXiv2020

Reanalysis of Variance Reduced Temporal Difference Learning

Temporal difference (TD) learning is a popular algorithm for policy evaluation in reinforcement learning, but the vanilla TD can substantially suffer from the inherent optimization variance. A variance reduced TD (VRTD) algorithm was proposed by Korda and La (2015), which applies the variance reduction technique directly to the online TD learning with Markovian samples. In this work, we first point out the technical errors in the analysis of VRTD in Korda and La (2015), and then provide a mathematically solid analysis of the non-asymptotic convergence of VRTD and its variance reduction performance. We show that VRTD is guaranteed to converge to a neighborhood of the fixed-point solution of TD at a linear convergence rate. Furthermore, the variance error (for both i.i.d.\ and Markovian sampling) and the bias error (for Markovian sampling) of VRTD are significantly reduced by the batch size of variance reduction in comparison to those of vanilla TD. As a result, the overall computational complexity of VRTD to attain a given accurate solution outperforms that of TD under Markov sampling and outperforms that of TD under i.i.d.\ sampling for a sufficiently small conditional number.

cs.LG↗

arXiv2019

HybridAlpha: An Efficient Approach for Privacy-Preserving Federated Learning

Federated learning has emerged as a promising approach for collaborative and privacy-preserving learning. Participants in a federated learning process cooperatively train a model by exchanging model parameters instead of the actual training data, which they might want to keep private. However, parameter interaction and the resulting model still might disclose information about the training data used. To address these privacy concerns, several approaches have been proposed based on differential privacy and secure multiparty computation (SMC), among others. They often result in large communication overhead and slow training time. In this paper, we propose HybridAlpha, an approach for privacy-preserving federated learning employing an SMC protocol based on functional encryption. This protocol is simple, efficient and resilient to participants dropping out. We evaluate our approach regarding the training time and data volume exchanged using a federated learning process to train a CNN on the MNIST data set. Evaluation against existing crypto-based SMC solutions shows that HybridAlpha can reduce the training time by 68% and data transfer volume by 92% on average while providing the same model performance and privacy guarantees as the existing solutions.

cs.CR↗

arXiv2019

A unified variance-reduced accelerated gradient method for convex optimization

We propose a novel randomized incremental gradient algorithm, namely, VAriance-Reduced Accelerated Gradient (Varag), for finite-sum optimization. Equipped with a unified step-size policy that adjusts itself to the value of the condition number, Varag exhibits the unified optimal rates of convergence for solving smooth convex finite-sum problems directly regardless of their strong convexity. Moreover, Varag is the first accelerated randomized incremental gradient method that benefits from the strong convexity of the data-fidelity term to achieve the optimal linear convergence. It also establishes an optimal linear rate of convergence for solving a wide class of problems only satisfying a certain error bound condition rather than strong convexity. Varag can also be extended to solve stochastic finite-sum problems.

math.OC↗

arXiv2019

Improved Zeroth-Order Variance Reduced Algorithms and Analysis for Nonconvex Optimization

Two types of zeroth-order stochastic algorithms have recently been designed for nonconvex optimization respectively based on the first-order techniques SVRG and SARAH/SPIDER. This paper addresses several important issues that are still open in these methods. First, all existing SVRG-type zeroth-order algorithms suffer from worse function query complexities than either zeroth-order gradient descent (ZO-GD) or stochastic gradient descent (ZO-SGD). In this paper, we propose a new algorithm ZO-SVRG-Coord-Rand and develop a new analysis for an existing ZO-SVRG-Coord algorithm proposed in Liu et al. 2018b, and show that both ZO-SVRG-Coord-Rand and ZO-SVRG-Coord (under our new analysis) outperform other exiting SVRG-type zeroth-order methods as well as ZO-GD and ZO-SGD. Second, the existing SPIDER-type algorithm SPIDER-SZO (Fang et al. 2018) has superior theoretical performance, but suffers from the generation of a large number of Gaussian random variables as well as a $\sqrtε$-level stepsize in practice. In this paper, we develop a new algorithm ZO-SPIDER-Coord, which is free from Gaussian variable generation and allows a large constant stepsize while maintaining the same convergence rate and query complexity, and we further show that ZO-SPIDER-Coord automatically achieves a linear convergence rate as the iterate enters into a local PL region without restart and algorithmic modification.

cs.LG↗

arXiv2019

On the exponential stability of a stratified flow to the 2D IDEAL MHD equations with damping

We study the stability of a type of stratified flows of the two dimensional inviscid incompressible MHD equations with velocity damping. The exponential stability for the perturbation near certain stratified flow is investigated in a strip-type area R*[0,1]. Although the magnetic filed potential is governed by a transport equation, by using the algebraic structure of the incompressible condition, it turns out that the linearized MHD equations around the given stratified flow retain a non-local damping mechanism. After carefully analyzing the non-linear structure and introducing some suitable weighted energy norms, we get the exponential stability by combining the exponential decay in time in the lower order energy with that in the high order energy.

math.AP↗

arXiv2019

Faster and Safer Training by Embedding High-Level Knowledge into Deep Reinforcement Learning

Deep reinforcement learning has been successfully used in many dynamic decision making domains, especially those with very large state spaces. However, it is also well-known that deep reinforcement learning can be very slow and resource intensive. The resulting system is often brittle and difficult to explain. In this paper, we attempt to address some of these problems by proposing a framework of Rule-interposing Learning (RIL) that embeds high level rules into the deep reinforcement learning. With some good rules, this framework not only can accelerate the learning process, but also keep it away from catastrophic explorations, thus making the system relatively stable even during the very early stage of training. Moreover, given the rules are high level and easy to interpret, they can be easily maintained, updated and shared with other similar tasks.

cs.AI↗

arXiv2019

Exotic Cooper pairing in multi-orbital models of Sr$_2$RuO$_4$

The unconventional superconductivity in Sr$_2$RuO$_4$ continues to defy a unified interpretation. In this paper, we focus on some novel aspects of its superconducting pairing by exploiting the orbital degree of freedom in this material. The multi-orbital nature, combined with the symmetry of the orbitals involved, leads to a plethora of exotic Cooper pairings not accessible in single-orbital systems. Essential physics is illustrated first using a two-orbital model with $d_{xz}$- and $d_{yz}$-orbitals. We classify the gap functions according to the underlying lattice symmetries, analyze the effective theories of a few representative pairings, and make connections to Sr$_2$RuO$_4$ in the course. In particular, we show how spin-orbit coupling may entangle spin-triplet and spin-singlet pairings. For completeness, the classification is generalized to the three-orbital model involving the $d_{xy}$-orbital as well. The orbital-basis approach distinguishes from the itinerant-band description for Sr$_2$RuO$_4$, and hence offers an alternative perspective to investigate its enigmatic superconducting state.

cond-mat.supr-con↗

arXiv2019

Supervised Encoding for Discrete Representation Learning

Classical supervised classification tasks search for a nonlinear mapping that maps each encoded feature directly to a probability mass over the labels. Such a learning framework typically lacks the intuition that encoded features from the same class tend to be similar and thus has little interpretability for the learned features. In this paper, we propose a novel supervised learning model named Supervised-Encoding Quantizer (SEQ). The SEQ applies a quantizer to cluster and classify the encoded features. We found that the quantizer provides an interpretable graph where each cluster in the graph represents a class of data samples that have a particular style. We also trained a decoder that can decode convex combinations of the encoded features from similar and different clusters and provide guidance on style transfer between sub-classes.

cs.CV↗

arXiv2019

Hybrid coarse-fine classification for head pose estimation

Head pose estimation, which computes the intrinsic Euler angles (yaw, pitch, roll) from the human, is crucial for gaze estimation, face alignment, and 3D reconstruction. Traditional approaches heavily relies on the accuracy of facial landmarks. It limits their performances, especially when the visibility of the face is not in good condition. In this paper, to do the estimation without facial landmarks, we combine the coarse and fine regression output together for a deep network. Utilizing more quantization units for the angles, a fine classifier is trained with the help of other auxiliary coarse units. Integrating regression is adopted to get the final prediction. The proposed approach is evaluated on three challenging benchmarks. It achieves the state-of-the-art on AFLW2000, BIWI and performs favorably on AFLW. The code has been released on Github.

cs.CV↗

arXiv2019

Towards Federated Graph Learning for Collaborative Financial Crimes Detection

Financial crime is a large and growing problem, in some way touching almost every financial institution. Financial institutions are the front line in the war against financial crime and accordingly, must devote substantial human and technology resources to this effort. Current processes to detect financial misconduct have limitations in their ability to effectively differentiate between malicious behavior and ordinary financial activity. These limitations tend to result in gross over-reporting of suspicious activity that necessitate time-intensive and costly manual review. Advances in technology used in this domain, including machine learning based approaches, can improve upon the effectiveness of financial institutions' existing processes, however, a key challenge that most financial institutions continue to face is that they address financial crimes in isolation without any insight from other firms. Where financial institutions address financial crimes through the lens of their own firm, perpetrators may devise sophisticated strategies that may span across institutions and geographies. Financial institutions continue to work relentlessly to advance their capabilities, forming partnerships across institutions to share insights, patterns and capabilities. These public-private partnerships are subject to stringent regulatory and data privacy requirements, thereby making it difficult to rely on traditional technology solutions. In this paper, we propose a methodology to share key information across institutions by using a federated graph learning platform that enables us to build more accurate machine learning models by leveraging federated learning and also graph learning approaches. We demonstrated that our federated model outperforms local model by 20% with the UK FCA TechSprint data set. This new platform opens up a door to efficiently detecting global money laundering activity.

cs.CY↗

arXiv2019

A Modularized Neural Network with Language-Specific Output Layers for Cross-lingual Voice Conversion

This paper presents a cross-lingual voice conversion framework that adopts a modularized neural network. The modularized neural network has a common input structure that is shared for both languages, and two separate output modules, one for each language. The idea is motivated by the fact that phonetic systems of languages are similar because humans share a common vocal production system, but acoustic renderings, such as prosody and phonotactic, vary a lot from language to language. The modularized neural network is trained to map Phonetic PosteriorGram (PPG) to acoustic features for multiple speakers. It is conditioned on a speaker i-vector to generate the desired target voice. We validated the idea between English and Mandarin languages in objective and subjective tests. In addition, mixed-lingual PPG derived from a unified English-Mandarin acoustic model is proposed to capture the linguistic information from both languages. It is found that our proposed modularized neural network significantly outperforms the baseline approaches in terms of speech quality and speaker individuality, and mixed-lingual PPG representation further improves the conversion performance.

eess.AS↗