Syed Fawad Hussain

Title Multi-view clustering: techniques and applications
Abstract

In many interesting domains, information about a particular instance can be gathered from different sources. Such instances are, therefore, represented by attributes that can naturally be split into multiple subsets. For example, a movie can be described either by its actors, its plot or a set of keywords used to describe the movie. Each set is referred to as a view and individually suffices for learning. In multi-view learning, such as clustering, we apply clustering on each of the individual view and use bootstrapping such that the final clustering gives better result than each of the individual view clustering. Multi-view clustering has recently been made popular and found application in several different areas including text analysis, social media, surveillance, etc. In this talk, we shall explore recent techniques that have been developed for multi-view clustering including extensions made to ensemble learning. In particular, I will discuss techniques for transferring information from one view to the other views in order to improve on the consensus result of the global view. I shall also discuss a very recent trend of using different dimensions of each view to further increase the performance of multi-view clustering and discuss application areas where multiview clustering have found success.

Bio

Syed Fawad Hussain obtained his M.S. in Computer science from Pierre and Marie Curie University, Paris, and a Ph.D. in Computer Science from the University of Grenoble, France. His research are during his Ph.D. was in Machine Learning during which he proposed new algorithms for finding similarity patterns in data. Dr. Fawad has published several research papers in leading conferences and journals. He is actively involved in many collaborative works, both locally and internationally and his research has resulted in several publications in leading international conferences and journals. He is also involved in other scholarly activities including being a reviewer for several international conferences and reputed ISI-indexed journals including those published by IEEE, Elsevier and Springer. Dr. Fawad is working as an assistant professor at the Faculty of Computer Sciences and Engineering, GIK Institute of Engineering Sciences and Technology with which he has been associated since 2010. His current area of research interest includes big data, unsupervised learning, similarity metrics, intelligent feature selection, information retrieval and bio-informatics.