Search

Your search keyword '"Liao, Xu"' showing total 1,111 results
Start Over Author "Liao, Xu"
1,111 results on '"Liao, Xu"'

11. Schr{\'o}dinger-F{\'o}llmer Sampler: Sampling without Ergodicity

Author

Huang, Jian, Jiao, Yuling, Kang, Lican, Liao, Xu, Liu, Jin, and Liu, Yanyan

Subjects
Statistics - Computation
Abstract

Sampling from probability distributions is an important problem in statistics and machine learning, specially in Bayesian inference when integration with respect to posterior distribution is intractable and sampling from the posterior is the only viable option for inference. In this paper, we propose Schr\"{o}dinger-F\"{o}llmer sampler (SFS), a novel approach for sampling from possibly unnormalized distributions. The proposed SFS is based on the Schr\"{o}dinger-F\"{o}llmer diffusion process on the unit interval with a time dependent drift term, which transports the degenerate distribution at time zero to the target distribution at time one. Comparing with the existing Markov chain Monte Carlo samplers that require ergodicity, no such requirement is needed for SFS. Computationally, SFS can be easily implemented using the Euler-Maruyama discretization. In theoretical analysis, we establish non-asymptotic error bounds for the sampling distribution of SFS in the Wasserstein distance under suitable conditions. We conduct numerical experiments to evaluate the performance of SFS and demonstrate that it is able to generate samples with better quality than several existing methods.

Published
2021

18. Deep Dimension Reduction for Supervised Representation Learning

Author

Huang, Jian, Jiao, Yuling, Liao, Xu, Liu, Jin, and Yu, Zhou

Subjects
Computer Science - Machine Learning, Statistics - Machine Learning
Abstract

The goal of supervised representation learning is to construct effective data representations for prediction. Among all the characteristics of an ideal nonparametric representation of high-dimensional complex data, sufficiency, low dimensionality and disentanglement are some of the most essential ones. We propose a deep dimension reduction approach to learning representations with these characteristics. The proposed approach is a nonparametric generalization of the sufficient dimension reduction method. We formulate the ideal representation learning task as that of finding a nonparametric representation that minimizes an objective function characterizing conditional independence and promoting disentanglement at the population level. We then estimate the target representation at the sample level nonparametrically using deep neural networks. We show that the estimated deep nonparametric representation is consistent in the sense that its excess risk converges to zero. Our extensive numerical experiments using simulated and real benchmark data demonstrate that the proposed methods have better performance than several existing dimension reduction methods and the standard deep learning models in the context of classification and regression.

Published
2020

23. Assessments of Pollution Status and Human Health Risk of Potentially Toxic Elements in Primary Crops and Agricultural Soils in Guanajuato, Mexico

Author

Liao, Xu, Li, Yanmei, Miranda-Avilés, Raúl, Puy-Alquiza, María Jesús, Bian, Jianmin, Anguiano, Jesús Horacio Hernández, Muñoz, Alma Hortensia Serafín, Datta, Saugata, Zha, Xiaoxiao, Liu, Jianli, Moncada, Daniela, Zhao, Zhiqing, González, Viridiana Piña, Garzón, Luisa Fernanda Rueda, Kshirsagar, Pooja, and Céspedes, Juan Manuel Navarro

Published
2023
Full Text
View/download PDF

28. Possible s-wave annihilation for MeV dark matter with the 21-cm absorption

Author

Jia, Lian-Bao and Liao, Xu

Subjects
High Energy Physics - Phenomenology, Astrophysics - Cosmology and Nongalactic Astrophysics
Abstract

The CMB observation sets stringent constraints on MeV dark matter (DM) annihilating into charged states/photons in s-wave, and the recent observation of the 21-cm absorption at the cosmic dawn reported by EDGES is also very strict for s-wave annihilations of MeV DM. The millicharged DM with p-wave dominant annihilations during the freeze-out period are considered in literatures to give an explanation about the 21-cm absorption, with photon mediated scattering cooling the hydrogen. In this paper, we focus on the annihilation of millicharged DM being s-wave dominant. To explain the 21-cm absorption and meanwhile be compatible with the CMB and 21-cm absorption bounds on DM annihilations, we consider the annihilation close to the resonance, with the new mediator (here is dark photon) mass being slightly above twice of the millicharged DM mass. In this case, the annihilation cross section at the temperature $T \to 0$ could be much smaller than that at $T_f$, which would be tolerated by the bounds on DM annihilations, avoiding the excess heating from DM s-wave annihilations to the hydrogen gas. The beam dump and lepton collider experiments can be employed to hunt for millicharged DM via the production of the invisible dark photon., Comment: 7 pages, 4 figures, discussions improved

Published
2019
Full Text
View/download PDF

46. Study on the Influence of HTS Strip and Harmonics on AC Loss of HTS Transformer

Author

Wang, Chen, Fang, Jin, Liao, Xu, Wang, Xiu, Zan, Yuanting, and Xiao, Zhaopeng

Abstract

The coil of the high-temperature superconducting (HTS) transformer is made of superconducting material, which has the characteristics of being lightweight, low loss, and having a low burning risk. However, when a superconductor operates or transmits alternating currents in an alternating magnetic field, the changing magnetic field will generate an inductive electric field, which will cause the AC loss of superconductors. On the one hand, the overall efficiency of the HTS transformer will be reduced, and on the other hand, the temperature of the HTS strip will increase, which will affect its critical current. When it is serious, it leads to the quench of the winding, accelerates the aging of the insulation layer, limits the overload capacity of the transformer, and even damages the entire transformer. In this paper, the influence of AC loss on HTS transformer winding is analyzed from two aspects of HTS strip and high-order harmonics. Based on the two-dimensional (2D) axisymmetrical coordinate finite element calculation model, the electromagnetic calculation and AC loss characteristic analysis of the 20 kVA superconducting traction transformer is carried out on the finite element simulation analysis software. Different HTS strips are adopted, and corresponding high-order harmonic currents are set. The influence of the HTS strip and high order harmonics on AC loss of the HTS transformer is obtained.

Published
2024
Full Text
View/download PDF

50. Therapies and Treatment Responses for Hemophagocytic Lymphohistiocytosis in Children: A Single-center Retrospective Study

Author

Yu, Lihua, primary, Lin, Danna, additional, Wu, Li, additional, Huang, Lulu, additional, Lai, Xiaorong, additional, Zhang, Yajie, additional, Zi, Juan, additional, Zhang, Jingxin, additional, Liao, Xu, additional, Liang, Lichan, additional, Zhang, Guanmei, additional, Yang, Liucheng, additional, and Yang, Lihua, additional

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results