arXiv subjects

Wei Sun

Publications and source records attributed to Wei Sun.

At least 415 records · Page 23Linked to original sources

arXiv2019

High-low level support vector regression prediction approach (HL-SVR) for data modeling with input parameters of unequal sample sizes

Support vector regression (SVR) has been widely used to reduce the high computational cost of computer simulation. SVR assumes the input parameters have equal sample sizes, but unequal sample sizes are often encountered in engineering practices. To solve this issue, a new prediction approach based on SVR, namely as high-low-level SVR approach (HL-SVR) is proposed for data modeling of input parameters of unequal sample sizes in this paper. The proposed approach is consisted of low-level SVR models for the input parameters of larger sample sizes and high-level SVR model for the input parameters of smaller sample sizes. For each training point of the input parameters of smaller sample sizes, one low-level SVR model is built based on its corresponding input parameters of larger sample sizes and their responses of interest. The high-level SVR model is built based on the obtained responses from the low-level SVR models and the input parameters of smaller sample sizes. Several numerical examples are used to validate the performance of HL-SVR. The experimental results indicate that HL-SVR can produce more accurate prediction results than conventional SVR. The proposed approach is applied on the stress analysis of dental implant, which the structural parameters have massive samples but the material of implant can only be selected from several Ti and its alloys. The prediction performance of the proposed approach is much better than the conventional SVR. The proposed approach can be used for the design, optimization and analysis of engineering systems with input parameters of unequal sample sizes.

eess.SP↗

arXiv2019

Observing topological charges and dynamical bulk-surface correspondence with ultracold atoms

In quenching a topological phase across phase transition, the dynamical bulk-surface correspondence emerges that the bulk topology of $d$-dimensional ($d$D) phase relates to the nontrivial pattern of quench dynamics emerging on $(d-1)$D subspace, called band inversion surfaces (BISs) in momentum space. Here we report the first experimental observation of the dynamical bulk-surface correspondence through measuring the topological charges in a 2D quantum anomalous Hall model realized in an optical Raman lattice. The system can be quenched with respect to every spin axis by suddenly varying the two-photon detuning or phases of the Raman couplings, in which the topological charges and BISs are measured dynamically by the time-averaged spin textures. We observe that the total charges in the region enclosed by BISs define a dynamical topological invariant, which equals the Chern index of the post-quench band. The topological charges relate to an emergent dynamical field which exhibits nontrivial topology on BIS, rendering the dynamical bulk-surface correspondence. This study opens a new avenue to explore topological phases dynamically.

cond-mat.quant-gas↗

arXiv2019

The Three-Dimensional Gaussian Product Inequality

We prove the 3-dimensional Gaussian product inequality, i.e., for any real-valued centered Gaussian random vector $(X,Y,Z)$ and $m\in \mathbb{N}$, it holds that ${\mathbf{E}}[X^{2m}Y^{2m}Z^{2m}]\geq{\mathbf{E}}[X^{2m}]{\mathbf{E}}[Y^{2m}]{\mathbf{E}}[Z^{2m}]$. Our proof is based on some improved inequalities on multi-term products involving 2-dimensional Gaussian random vectors. The improved inequalities are derived using the Gaussian hypergeometric functions and have independent interest. As by-products, several new combinatorial identities and inequalities are obtained.

math.PR↗

arXiv2019

Observation of the Hopf Links and Hopf Fibration in a 2D topological Raman Lattice

A dynamical Hopf insulator is experimentally synthesized with a quenched two-dimensional quantum anomalous Hall system on a square Raman lattice. The quench dynamics for the quasimomentum-time-dependent Bloch vectors defines a Hopf map from $(q_x,q_y,t)\in T^3$ to the Bloch sphere $S^2$. In this Hopf map, a dynamical Hopf number can be defined, and it exactly equals the Chern number of the post-quench Hamiltonian. We experimentally measure the Hopf link between the fibers for the North and South Poles on $S^2$, which are the trajectories in $(q_x,q_y,t)$ space with maximal spin polarization, to extract the topological Chern number of the post-quench Hamiltonian. We also observe the structure of Hopf fibration for the mutually nested Hopf tori. Our study sheds some new light on the interplay between topology (Hopf number) and geometry (fiber bundle) in quantum dynamics.

cond-mat.quant-gas↗

arXiv2019

Dynamic Learning with Frequent New Product Launches: A Sequential Multinomial Logit Bandit Problem

Motivated by the phenomenon that companies introduce new products to keep abreast with customers' rapidly changing tastes, we consider a novel online learning setting where a profit-maximizing seller needs to learn customers' preferences through offering recommendations, which may contain existing products and new products that are launched in the middle of a selling period. We propose a sequential multinomial logit (SMNL) model to characterize customers' behavior when product recommendations are presented in tiers. For the offline version with known customers' preferences, we propose a polynomial-time algorithm and characterize the properties of the optimal tiered product recommendation. For the online problem, we propose a learning algorithm and quantify its regret bound. Moreover, we extend the setting to incorporate a constraint which ensures every new product is learned to a given accuracy. Our results demonstrate the tier structure can be used to mitigate the risks associated with learning new products.

cs.LG↗

arXiv2019

Ultra-low noise magnetic field for quantum gases

Ultra-low noise magnetic field is essential for many branches of scientific research. Examplesinclude experiments conducted on ultra-cold atoms, quantum simulations, as well as precisionmeasurements. In ultra-cold atom experiments specifically, a bias magnetic field will be oftenserved as a quantization axis and be applied for Zeeman splitting. As atomic states areusually sensitive to magnetic fields, a magnetic field characterized by ultra-low noise as wellas high stability is typically required for experimentation. For this study, a bias magneticfield is successfully stabilized at 14.5G, with the root mean square (RMS) value of the noisereduced to 18.5μG (1.28ppm) by placingμ-metal magnetic shields together with a dynamicalfeedback circuit. Long-time instability is also regulated consistently below 7μG. The level ofnoise exhibited in the bias magnetic field is further confirmed by evaluating the coherencetime of a Bose-Einstein condensate characterized by Rabi oscillation. It is concluded thatthis approach can be applied to other physical systems as well.

physics.app-ph↗

arXiv2019

Two Theorems on Hunt's Hypothesis (H) for Markov Processes

Hunt's hypothesis (H) and the related Getoor's conjecture is one of the most important problems in the basic theory of Markov processes. In this paper, we investigate the invariance of Hunt's hypothesis (H) for Markov processes under two classes of transformations, which are change of measure and subordination. Our first theorem shows that for two standard processes $(X_t)$ and $(Y_t)$, if $(X_t)$ satisfies (H) and $(Y_t)$ is locally absolutely continuous with respect to $(X_t)$, then $(Y_t)$ satisfies (H). Our second theorem shows that a standard process $(X_t)$ satisfies (H) if and only if $(X_{τ_t})$ satisfies (H) for some (and hence any) subordinator $(τ_t)$ which is independent of $(X_t)$ and has a positive drift coefficient. Applications of the two theorems are given.

math.PR↗

arXiv2019

Ambient Electro-Synthesis of Ammonia - Electrode Porosity and Composition Engineering

Ammonia, key precursor for fertilizer production, convenient hydrogen carrier and emerging clean fuel, plays a pivotal role in sustaining life on earth. Currently, the main route for NH3 synthesis is via the heterogeneous catalytic Haber-Bosch process (N2+3H2 - 2NH3), which proceeds under extreme conditions of temperature and pressure with a very large carbon footprint. Herein we report that a pristine nitrogen-doped nanoporous graphitic carbon membrane (NCM) can electrochemically convert N2 into NH3 in an aqueous acidic solution under ambient conditions. The Faradaic efficiency and rate of production of NH3 on the NCM electrode reach 5.2% and 0.08 g m-2 h-1, respectively. After functionalization of the NCM with Au nanoparticles (Au NPs) these performance metrics are dramatically enhanced to 22% and 0.36 g m-2 h-1, respectively. These efficiencies and rates for the production of NH3 at room temperature and atmospheric pressure are unprecedented. As this system offers the potential to be scaled to industrial proportions there is a high likelihood it might displace the century old Haber-Bosch process.

physics.chem-ph↗

arXiv2019

Ranking-Based Autoencoder for Extreme Multi-label Classification

Extreme Multi-label classification (XML) is an important yet challenging machine learning task, that assigns to each instance its most relevant candidate labels from an extremely large label collection, where the numbers of labels, features and instances could be thousands or millions. XML is more and more on demand in the Internet industries, accompanied with the increasing business scale / scope and data accumulation. The extremely large label collections yield challenges such as computational complexity, inter-label dependency and noisy labeling. Many methods have been proposed to tackle these challenges, based on different mathematical formulations. In this paper, we propose a deep learning XML method, with a word-vector-based self-attention, followed by a ranking-based AutoEncoder architecture. The proposed method has three major advantages: 1) the autoencoder simultaneously considers the inter-label dependencies and the feature-label dependencies, by projecting labels and features onto a common embedding space; 2) the ranking loss not only improves the training efficiency and accuracy but also can be extended to handle noisy labeled data; 3) the efficient attention mechanism improves feature representation by highlighting feature importance. Experimental results on benchmark datasets show the proposed method is competitive to state-of-the-art methods.

cs.LG↗

arXiv2019

Dynamic Learning of Sequential Choice Bandit Problem under Marketing Fatigue

Motivated by the observation that overexposure to unwanted marketing activities leads to customer dissatisfaction, we consider a setting where a platform offers a sequence of messages to its users and is penalized when users abandon the platform due to marketing fatigue. We propose a novel sequential choice model to capture multiple interactions taking place between the platform and its user: Upon receiving a message, a user decides on one of the three actions: accept the message, skip and receive the next message, or abandon the platform. Based on user feedback, the platform dynamically learns users' abandonment distribution and their valuations of messages to determine the length of the sequence and the order of the messages, while maximizing the cumulative payoff over a horizon of length T. We refer to this online learning task as the sequential choice bandit problem. For the offline combinatorial optimization problem, we show that an efficient polynomial-time algorithm exists. For the online problem, we propose an algorithm that balances exploration and exploitation, and characterize its regret bound. Lastly, we demonstrate how to extend the model with user contexts to incorporate personalization.

cs.LG↗

arXiv2019

Real time backbone for semantic segmentation

The rapid development of autonomous driving in recent years presents lots of challenges for scene understanding. As an essential step towards scene understanding, semantic segmentation thus received lots of attention in past few years. Although deep learning based state-of-the-arts have achieved great success in improving the segmentation accuracy, most of them suffer from an inefficiency problem and can hardly applied to practical applications. In this paper, we systematically analyze the computation cost of Convolutional Neural Network(CNN) and found that the inefficiency of CNN is mainly caused by its wide structure rather than the deep structure. In addition, the success of pruning based model compression methods proved that there are many redundant channels in CNN. Thus, we designed a very narrow while deep backbone network to improve the efficiency of semantic segmentation. By casting our network to FCN32 segmentation architecture, the basic structure of most segmentation methods, we achieved 60.6\% mIoU on Cityscape val dataset with 54 frame per seconds(FPS) on $1024\times2048$ inputs, which already outperforms one of the earliest real time deep learning based segmentation methods: ENet.

cs.CV↗

arXiv2019

OpenFermion: The Electronic Structure Package for Quantum Computers

Quantum simulation of chemistry and materials is predicted to be an important application for both near-term and fault-tolerant quantum devices. However, at present, developing and studying algorithms for these problems can be difficult due to the prohibitive amount of domain knowledge required in both the area of chemistry and quantum algorithms. To help bridge this gap and open the field to more researchers, we have developed the OpenFermion software package (www.openfermion.org). OpenFermion is an open-source software library written largely in Python under an Apache 2.0 license, aimed at enabling the simulation of fermionic models and quantum chemistry problems on quantum hardware. Beginning with an interface to common electronic structure packages, it simplifies the translation between a molecular specification and a quantum circuit for solving or studying the electronic structure problem on a quantum computer, minimizing the amount of domain expertise required to enter the field. The package is designed to be extensible and robust, maintaining high software standards in documentation and testing. This release paper outlines the key motivations behind design choices in OpenFermion and discusses some basic OpenFermion functionality which we believe will aid the community in the development of better quantum algorithms and tools for this exciting area of research.

quant-ph↗

arXiv2019

Learning Spatial Pyramid Attentive Pooling in Image Synthesis and Image-to-Image Translation

Image synthesis and image-to-image translation are two important generative learning tasks. Remarkable progress has been made by learning Generative Adversarial Networks (GANs)~\cite{goodfellow2014generative} and cycle-consistent GANs (CycleGANs)~\cite{zhu2017unpaired} respectively. This paper presents a method of learning Spatial Pyramid Attentive Pooling (SPAP) which is a novel architectural unit and can be easily integrated into both generators and discriminators in GANs and CycleGANs. The proposed SPAP integrates Atrous spatial pyramid~\cite{chen2018deeplab}, a proposed cascade attention mechanism and residual connections~\cite{he2016deep}. It leverages the advantages of the three components to facilitate effective end-to-end generative learning: (i) the capability of fusing multi-scale information by ASPP; (ii) the capability of capturing relative importance between both spatial locations (especially multi-scale context) or feature channels by attention; (iii) the capability of preserving information and enhancing optimization feasibility by residual connections. Coarse-to-fine and fine-to-coarse SPAP are studied and intriguing attention maps are observed in both tasks. In experiments, the proposed SPAP is tested in GANs on the Celeba-HQ-128 dataset~\cite{karras2017progressive}, and tested in CycleGANs on the Image-to-Image translation datasets including the Cityscape dataset~\cite{cordts2016cityscapes}, Facade and Aerial Maps dataset~\cite{zhu2017unpaired}, both obtaining better performance.

cs.CV↗

arXiv2018

High Controllable and Robust 2D Spin-Orbit Coupling for Quantum Gases

We report the realization of a robust and highly controllable two-dimensional (2D) spin-orbit (SO) coupling with topological non-trivial band structure. By applying a retro-reflected 2D optical lattice, phase tunable Raman couplings are formed into the anti-symmetric Raman lattice structure, and generate the 2D SO coupling with precise inversion and $C_4$ symmetries, leading to considerably enlarged topological regions. The life time of the 2D SO coupled Bose-Einstein condensate reaches several seconds, which enables the exploring of fine tuning interaction effects. These essential advantages of the present new realization open the door to explore exotic quantum many-body effects and non-equilibrium dynamics with novel topology.

cond-mat.quant-gas↗

arXiv2018

Glueball spectrum from $N_f=2$ lattice QCD study on anisotropic lattices

The lowest-lying glueballs are investigated in lattice QCD using $N_f=2$ clover Wilson fermion on anisotropic lattices. We simulate at two different and relatively heavy quark masses, corresponding to physical pion mass of $m_π\sim 938$ MeV and $650$ MeV. The quark mass dependence of the glueball masses have not been investigated in the present study. Only the gluonic operators built from Wilson loops are utilized in calculating the corresponding correlation functions. In the tensor channel, we obtain the ground state mass to be 2.363(39) GeV and 2.384(67) GeV at $m_π\sim 938$ MeV and $650$ MeV, respectively. In the pseudoscalar channel, when using the gluonic operator whose continuum limit has the form of $ε_{ijk}TrB_iD_jB_k$, we obtain the ground state mass to be 2.573(55) GeV and 2.585(65) GeV at the two pion masses. These results are compatible with the corresponding results in the quenched approximation. In contrast, if we use the topological charge density as field operators for the pseudoscalar, the masses of the lowest state are much lighter (around 1GeV) and compatible with the expected masses of the flavor singlet $q\bar{q}$ meson. This indicates that the operator $ε_{ijk}TrB_iD_jB_k$ and the topological charge density couple rather differently to the glueball states and $q\bar{q}$ mesons. The observation of the light flavor singlet pseudoscalar meson can be viewed as the manifestation of effects of dynamical quarks. In the scalar channel, the ground state masses extracted from the correlation functions of gluonic operators are determined to be around 1.4-1.5 GeV, which is close to the ground state masses from the correlation functions of the quark bilinear operators. In all cases, the mixing between glueballs and conventional mesons remains to be further clarified in the future.

hep-lat↗

arXiv2018

Geology prediction based on operation data of TBM: comparison between deep neural network and statistical learning methods

Tunnel boring machine (TBM) is a complex engineering system widely used for tunnel construction. In view of the complicated construction environments, it is necessary to predict geology conditions prior to excavation. In recent years, massive operation data of TBM has been recorded, and mining these data can provide important references and useful information for designers and operators of TBM. In this work, a geology prediction approach is proposed based on deep neural network and operation data. It can provide relatively accurate geology prediction results ahead of the tunnel face compared with the other prediction models based on statistical learning methods. The application case study on a tunnel in China shows that the proposed approach can accurately estimate the geological conditions prior to excavation, especially for the short range ahead of training data. This work can be regarded as a good complement to the geophysical prospecting approach during the construction of tunnels, and also highlights the applicability and potential of deep neural networks for other data mining tasks of TBMs.

physics.geo-ph↗

arXiv2018

Towards Interpretable R-CNN by Unfolding Latent Structures

This paper first proposes a method of formulating model interpretability in visual understanding tasks based on the idea of unfolding latent structures. It then presents a case study in object detection using popular two-stage region-based convolutional network (i.e., R-CNN) detection systems. We focus on weakly-supervised extractive rationale generation, that is learning to unfold latent discriminative part configurations of object instances automatically and simultaneously in detection without using any supervision for part configurations. We utilize a top-down hierarchical and compositional grammar model embedded in a directed acyclic AND-OR Graph (AOG) to explore and unfold the space of latent part configurations of regions of interest (RoIs). We propose an AOGParsing operator to substitute the RoIPooling operator widely used in R-CNN. In detection, a bounding box is interpreted by the best parse tree derived from the AOG on-the-fly, which is treated as the qualitatively extractive rationale generated for interpreting detection. We propose a folding-unfolding method to train the AOG and convolutional networks end-to-end. In experiments, we build on R-FCN and test our method on the PASCAL VOC 2007 and 2012 datasets. We show that the method can unfold promising latent structures without hurting the performance.

cs.CV↗

arXiv2018

Gaussian process regression for survival time prediction with genome-wide gene expression

Predicting the survival time of a cancer patient based on his/her genome-wide gene expression remains a challenging problem. For certain types of cancer, the effects of gene expression on survival are both weak and abundant, so identifying nonzero effects with reasonable accuracy is difficult. As an alternative to methods that use variable selection, we propose a Gaussian process accelerated failure time model to predict survival time using genome-wide or pathway-wide gene expression data. Using a Monte Carlo EM algorithm, we jointly impute censored log-survival time and estimate model parameters. We demonstrate the performance of our method and its advantage over existing methods in both simulations and real data analysis. The real data that we analyze were collected from 513 patients with kidney renal clear cell carcinoma and include survival time, demographic/clinical variables, and expression of more than 20,000 genes. Our method is widely applicable as it can accommodate right, left, and interval censored outcomes; and provides a natural way to combine multiple types of high-dimensional -omics data. An R package implementing our method is available online.

stat.ME↗