A latent Beta-Liouville allocation model

作者：

Highlights：

• A latent Beta-Liouville allocation model is proposed.

• The proposed model is learned using a principled variational approach.

• The model is applied to the challenging problems of visual scene and text categorization, and action recognition.

摘要

•A latent Beta-Liouville allocation model is proposed.•The proposed model is learned using a principled variational approach.•The model is applied to the challenging problems of visual scene and text categorization, and action recognition.

论文关键词：Latent topic models,Count data,Beta-Liouville distribution,Variational Bayes,Graphical models,Text classification,Scene categorization,Action recognition

论文评审过程：Received 29 April 2014, Revised 1 September 2015, Accepted 27 September 2015, Available online 9 October 2015, Version of Record 10 November 2015.

论文官网地址：https://doi.org/10.1016/j.eswa.2015.09.044