Learning Multi Labels from Single Label ---- An Extreme Weak Label Learning Algorithm

Citation metadata

Date: Apr. 2019
Publisher: Springer
Document Type: Report
Length: 170 words

Document controls

Main content

Abstract :

Byline: Junhong Duan (1), Xiaoyu Li (1), Dejun Mu (1) Keywords: weak-supervised learning; genetic algorithm; multi-label classification; TP 393 Abstract: This paper presents a novel algorithm for an extreme form of weak label learning, in which only one of all relevant labels is given for each training sample. Using genetic algorithm, all of the labels in the training set are optimally divided into several non-overlapping groups to maximize the label distinguishability in every group. Multiple classifiers are trained separately and ensembled for label predictions. Experimental results show significant improvement over previous weak label learning algorithms. Author Affiliation: (1) 0000 0001 0307 1240, grid.440588.5, Shenzhen Research Institute, Northwestern Polytechnical University, Shenzhen, Guangdong, 518057, China Article History: Registration Date: 19/03/2019 Received Date: 20/07/2018 Online Date: 20/03/2019 Article note: Foundation item: Supported by the National Natural Science Foundation of China (61672433), the Fundamental Research Fund for Shenzhen Science and Technology Innovation Committee (201703063000511, 201703063000517), the National Cryptography Development Fund (MMJJ20170210) and the Science and Technology Project of State Grid Corporation of China (522722180007)

Source Citation

Source Citation   

Gale Document Number: GALE|A579551807