主讲人介绍：

Zhirong Yang received his Bachelor and Master degrees from Sun Yat-Sen University, China in 1999 and 2002, and Doctoral degree from Helsinki University of Technology, Finland in 2008, respectively. Presently he is a professor with the Norwegian Open AI Lab and Department of Computer Science at Norwegian University of Science and Technology. He also works as Finnish Academy Research Fellow at Aalto University, Finland. His current research interests include information visualization, cluster analysis, learning representations for structured data.

内容摘要：

Unsupervised learning has been one of the focuses in recent machine learning research. In this talk I will briefly review the progress in two fundamental unsupervised learning tasks: cluster analysis and nonlinear dimensionality reduction for data visualization. Our work is based on information divergence and manifold learning, which provides a scalable framework for large-scale and sparse similarities. On top of findings in individual tasks, we will also discuss a new direction which combines the ideas in the two research lines.