Stochastic variational inference is a powerful tool for analyzing probabilistic models, especially for large scale problem. In this project, our goal is to implement a generic algorithm based on stochastic variational inference for a large family of probabilistic models, including latent Dirichlet allocation (LDA), hidden Markov models (HMM), and Bayesian mixture models (BMM) and etc. We describe the design and the implementation plan of our project and propose a timeline for our development.

Student

Zhehui Chen

Mentors

  • Xingguo
  • Tuo Zhao
close

2017