Robust Boosting Algorithm Against Mislabeling in Multiclass Problems
Robust Boosting Algorithm Against Mislabeling in Multiclass Problems
复制标题
DOI：
10.1162/neco.2007.11-06-400
复制
发表时间：
2008-06
影响因子：
2.9
通讯作者：
Takashi Takenouchi;S. Eguchi;Noboru Murata;T. Kanamori
中科院分区：
文献类型：
--
作者：
Takashi Takenouchi;S. Eguchi;Noboru Murata;T. Kanamori
我们讨论了在多类标签分类问题中对错误标记的鲁棒性，并提出了两种基于Eta-divergence的boosting算法，归一化的Eta-Boost.M和Eta-Boost.M。这两种增强算法与错误标签模型密切相关，其中标签被错误地交换为其他标签。对于这两种增强算法，探索了支持错误标记鲁棒性的理论方面。我们应用所提出的两种提升方法的合成和真实的数据集调查这些方法的性能，重点是鲁棒性，并确认所提出的方法的有效性。
We discuss robustness against mislabeling in multiclass labels for classification problems and propose two algorithms of boosting, the normalized Eta-Boost.M and Eta-Boost.M, based on the Eta-divergence. Those two boosting algorithms are closely related to models of mislabeling in which the label is erroneously exchanged for others. For the two boosting algorithms, theoretical aspects supporting the robustness for mislabeling are explored. We apply the proposed two boosting methods for synthetic and real data sets to investigate the performance of these methods, focusing on robustness, and confirm the validity of the proposed methods.