Analyzing Co-training Style Algorithms

Co-training is a semi-supervised learning paradigm which trains two learners respectively from two different views and lets the learners label some unlabeled examples for each other. In this paper, we present a new PAC analysis on co-training style algorithms. We show that the co-training process can succeed even without two views, given that the two learners have large difference, which explains the success of some co-training style algorithms that do not require two views. Moreover, we theoretically explain that why the co-training process could not improve the performance further after a number of rounds, and present a rough estimation on the appropriate round to terminate co-training to avoid some wasteful learning rounds.

• Type: Conference
• Year: 2007
• Where: ECML
• Authors: Wei Wang, Zhi-Hua Zhou
• Keywords: Co-training, Co-training Process, Co-training Style Algorithms, ECML 2007, Machine Learning