https://scholars.lib.ntu.edu.tw/handle/123456789/634870
標題: | Reduction from Complementary-Label Learning to Probability Estimates | 作者: | Lin, Wei I. HSUAN-TIEN LIN |
關鍵字: | complementary-label learning | weakly-supervised learning | 公開日期: | 1-一月-2023 | 卷: | 13936 LNCS | 來源出版物: | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | 摘要: | Complementary-Label Learning (CLL) is a weakly-supervised learning problem that aims to learn a multi-class classifier from only complementary labels, which indicate a class to which an instance does not belong. Existing approaches mainly adopt the paradigm of reduction to ordinary classification, which applies specific transformations and surrogate losses to connect CLL back to ordinary classification. Those approaches, however, face several limitations, such as the tendency to overfit. In this paper, we sidestep those limitations with a novel perspective–reduction to probability estimates of complementary classes. We prove that accurate probability estimates of complementary labels lead to good classifiers through a simple decoding step. The proof establishes a reduction framework from CLL to probability estimates. The framework offers explanations of several key CLL approaches as its special cases and allows us to design an improved algorithm that is more robust in noisy environments. The framework also suggests a validation procedure based on the quality of probability estimates, offering a way to validate models with only CLs. The flexible framework opens a wide range of unexplored opportunities in using deep and non-deep models for probability estimates to solve CLL. Empirical experiments further verified the framework’s efficacy and robustness in various settings. The full paper can be accessed at https://arxiv.org/abs/2209.09500. |
URI: | https://scholars.lib.ntu.edu.tw/handle/123456789/634870 | ISBN: | 9783031333767 | ISSN: | 03029743 | DOI: | 10.1007/978-3-031-33377-4_36 |
顯示於: | 資訊工程學系 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。