Fast sampling methods for Bayesian max-margin models
详细信息 查看全文
- 作者：Wenbo Hu hwb13@mails.tsinghua.edu.cn" class="auth_mail" title="E-mail the corresponding author ; Jun Zhu ; dcszj@mail.tsinghua.edu.cn" class="auth_mail" title="E-mail the corresponding author ; Bo Zhang dcszb@mail.tsinghua.edu.cn" class="auth_mail" title="E-mail the corresponding author
- 关键词：Inference ; Stochastic MCMC ; Subgradient MCMC ; Bayesian max-margin models ; Approximate detailed balance
- 刊名：Expert Systems with Applications
- 出版年：2017
- 出版时间：1 March 2017
- 年：2017
- 卷：69
- 期：Complete
- 页码：277-287
- 全文大小：973 K
文摘
- class="listitem" id="list_celist0004">
- class="label">•
We study the stochastic subgradient MCMC methods for Bayesian max-margin models.
- class="label">•
Theoretical analysis shows the approximate detailed balance property of our methods.
- class="label">•
Stochastic subsampling and thermostats are used for fast convergence and mixing.
- class="label">•
Experiments show the efficiency and accuracy of our methods.