CATVI: conditional and adaptively truncated variational inference for hierarchical Bayesian nonparametric models

Qiao, Xinghao ORCID: 0000-0002-6546-6595, Liu, Yirui and Lam, Jessica (2022) CATVI: conditional and adaptively truncated variational inference for hierarchical Bayesian nonparametric models. Proceedings of Machine Learning Research, 151. ISSN 2640-3498

Text (CATVI. Conditional and Adaptively Truncated Variational) - Accepted Version Download (624kB)

• Author
• Publisher
• https://proceedings.mlr.press/v151/liu22...

Abstract

Current variational inference methods for hierarchical Bayesian nonparametric models can neither characterize the correlation struc- ture among latent variables due to the mean- eld setting, nor infer the true posterior dimension because of the universal trunca- tion. To overcome these limitations, we pro- pose the conditional and adaptively trun- cated variational inference method (CATVI) by maximizing the nonparametric evidence lower bound and integrating Monte Carlo into the variational inference framework. CATVI enjoys several advantages over tra- ditional methods, including a smaller diver- gence between variational and true posteri- ors, reduced risk of undertting or overt- ting, and improved prediction accuracy. Em- pirical studies on three large datasets re- veal that CATVI applied in Bayesian non- parametric topic models substantially out- performs competing models, providing lower perplexity and clearer topic-words clustering.

Item Type:	Article
Official URL:	http://aistats.org/aistats2022/
Additional Information:	© 2022 The Authors
Divisions:	Statistics
Subjects:	H Social Sciences > HA Statistics Q Science > QA Mathematics
Date Deposited:	08 Apr 2022 15:48
Last Modified:	02 Oct 2024 15:45
URI:	http://eprints.lse.ac.uk/id/eprint/114639

Actions (login required)

View Item