Guest Talk: Kai Zhang, Large Scale Unsupervised and Semi-Supervised Learning
Kai Zhang (Computer & Information Sciences, Temple University),
Topic: Matrix approximation.
Nov 28, 3pm, SBA
Title: Large Scale Unsupervised and Semi-Supervised Learning
Abstract
Clustering is a fundamental data explorative step in pattern recognition and machine learning. This talk involves two types of clustering paradigms, the mixture models and graph-based clustering methods, with the primary focus on how to improve the scaling behavior of related algorithms for large-scale application. The first part is on simplifying mixture models, and the second part is on applying low-rank matrix approximation with novel sampling scheme in a large family of mainstream learning algorithms. Topics on community detection and large scale supervised learning (SVM) will also be included.