Fast copula variational inference

doi:10.1080/0952813X.2021.1871970

CIOMP OpenIR

	Fast copula variational inference
	J. Chi; J. Ouyang; A. Zhang; X. Wang and X. Li
	2022
发表期刊	Journal of Experimental and Theoretical Artificial Intelligence
ISSN	0952813X
卷号	34 期号:2 页码:295-310
摘要	Mean-field variational inference, built on fully factorisations, can be efficiently solved; however, it ignores the dependencies between latent variables, resulting in lower performance. To address this, the copula variational inference (CVI) method is proposed by using the well-established copulas to effectively capture posterior dependencies, leading to better approximations. However, it suffers from a computational issue, where the optimisation for big models with massive latent variables is quite time-consuming. This is mainly caused by the expensive sampling when forming noisy Monte Carlo gradients in CVI. For CVI speedup, in this paper we propose a novel fast CVI (abbr. FCVI). In FCVI, we derive the gradient of CVI objective by an expectation of the mean-field factorisation. Therefore, we can achieve a much efficient sampling from the (Formula presented.) -dimensional mean-field factorisation, enabling to reduce the sampling complexity from (Formula presented.) to (Formula presented.). To evaluate FCVI, we compare it against baseline methods on modelling performance and runtime. Experimental results demonstrate that FCVI is on a par with CVI, but runs much faster. 2021 Informa UK Limited, trading as Taylor & Francis Group.
DOI	10.1080/0952813X.2021.1871970
URL	查看原文
收录类别	ei
引用统计
文献类型	期刊论文
条目标识符	http://ir.ciomp.ac.cn/handle/181722/66663
专题	中国科学院长春光学精密机械与物理研究所
推荐引用方式 GB/T 7714	J. Chi,J. Ouyang,A. Zhang,et al. Fast copula variational inference[J]. Journal of Experimental and Theoretical Artificial Intelligence,2022,34(2):295-310.
APA	J. Chi,J. Ouyang,A. Zhang,&X. Wang and X. Li.(2022).Fast copula variational inference.Journal of Experimental and Theoretical Artificial Intelligence,34(2),295-310.
MLA	J. Chi,et al."Fast copula variational inference".Journal of Experimental and Theoretical Artificial Intelligence 34.2(2022):295-310.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
Fast copula variatio（3614KB）	期刊论文	出版稿	开放获取	CC BY-NC-SA	浏览下载