PAC-Bayes analysis of multi-view learning
详细信息 查看全文
- 作者：Shiliang Sun ; a ; shiliangsun@gmail.com" class="auth_mail" title="E-mail the corresponding author ; slsun@cs.ecnu.edu.cn" class="auth_mail" title="E-mail the corresponding author ; John Shawe-Taylorb ; Liang Maoa
- 关键词：PAC-Bayes bound ; Statistical learning theory ; Support vector machine ; Multi-view learning
- 刊名：Information Fusion
- 出版年：2017
- 出版时间：May 2017
- 年：2017
- 卷：35
- 期：Complete
- 页码：117-131
- 全文大小：634 K
文摘
- class="listitem" id="list_celist0001">
- class="label">•
Multi-view learning is a promising research direction with wide applicability.
- class="label">•
We present eight PAC-Bayes bounds to analyze the theoretical performance.
- class="label">•
Data dependent Gaussian priors are adopted.
- class="label">•
The bounds are evaluated and compared with each other.
- class="label">•
The usefulness and performance of the bounds are discussed.