Location and Venue: Tuesday, 2.9.2013, 11.30-12.30 in the Mathematics and Computer Science Department of University of Münster.
Learning in the Model Space for Temporal Data
Prof. Dr. Peter Tino
(University of Birmingham, UK)
I will first introduce the concept of learning in the model space. This talk will focus on time series data. After reviewing recent developments in model based time series kernels, I will introduce a framework for building new kernels based on temporal filters inspired by a class of "reservoir" models known as Echo State Networks. I will briefly outline the key theoretical concepts of their analysis and design. The methodology will be demonstrated in a series of sequence classification tasks and in an incremental fault detection setting.
