首页> 外文会议>International symposium on intelligent data analysis >Multiview Learning of Weighted Majority Vote by Bregman Divergence Minimization
【24h】

Multiview Learning of Weighted Majority Vote by Bregman Divergence Minimization

机译:通过Bregman发散最小化进行加权多数投票的多视图学习。

获取原文

摘要

We tackle the issue of classifier combinations when observations have multiple views. Our method jointly learns view-specific weighted majority vote classifiers (i.e. for each view) over a set of base voters, and a second weighted majority vote classifier over the set of these view-specific weighted majority vote classifiers. We show that the empirical risk minimization of the final majority vote given a multiview training set can be cast as the minimization of Bregman divergences. This allows us to derive a parallel-update optimization algorithm for learning our multiview model. We empirically study our algorithm with a particular focus on the impact of the training set size on the multi-view learning results. The experiments show that our approach is able to overcome the lack of labeled information.
机译:当观察结果具有多个视图时,我们将解决分类器组合的问题。我们的方法联合学习一组基础投票者的特定于视图的加权多数投票分类器(即针对每个视图),以及针对这些特定于视图的加权多数投票分类器的集合学习第二加权多数投票分类器。我们表明,给定多视图培训集,将最终多数票的经验风险最小化可以看作是Bregman差异的最小化。这使我们能够导出并行更新优化算法,以学习我们的多视图模型。我们通过经验研究我们的算法,特别关注训练集大小对多视图学习结果的影响。实验表明,我们的方法能够克服标签信息的不足。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号