Usual time: Thursdays 14:00-15:00
Location: Room 102, Department of Statistical Science, 1-19 Torrington Place (1st floor). Some seminars are held at different locations and at different times. Please click on the abstract for further details.
- 19 July 2018: Prof. Baogang Hu (Chinese Academy of Sciences)
Information theoretic learning (ITL) in pattern classification
When ITL (termed by Principe, et al. 2000) criteria have played more roles in the study of machine learning, we are still puzzled by their successes and limitations in applications, or their connections to the empirical learning criteria. This talk will focus on ITL in the study of pattern classification. The connections between the two sets of learning criteria are presented in a binary classification. I will introduce the novel theory of abstaining learning for both Bayesian classifiers and mutual information classifiers, and the cost-free learning from the real-world data sets in comparison with cost-sensitive learning. The talk will show that ITL will provide us a fundamental for understanding the learning mechanisms.