Approximation Analysis of some Learning Algorithms

Ding-Xuan Zhou (City University of Hong Kong)

Frank Adams 2,

Machine learning and learning algorithms aim at learning function relations or data structures from samples and are widely applied in data analysis. Learning theory provides rigorous mathematical foundations of machine learning, where approximation theory is important to understand learning ability of learning algorithms generated in some hypothesis spaces. In this talk we will briefly describe approximation analysis of some learning algorithms. Least squares regularized regression, support vector machine classification, and minimum error entropy principle will be demonstrated.

Import this event to your Outlook calendar
▲ Up to the top