Presenter: Prof Jian Pei, Simon Fraser University, Canada (invited by Prof Longbing Cao)
Abstract: Uncertainty is inherent and ubiquitous in many applications. In this talk, I will present our recent explorations on analyzing large uncertain data sets on several interesting aspects. First, to understand and summarize an uncertain object, I will illustrate a method to pick the top-k typical instances for an uncertain object. Second, I will elaborate how spatial analysis methods like skyline and nearest neighbor search can be extended to uncertain data, and, more importantly, what may become tricky. Last, I will exemplify the pros and cons of using the recently prevailing possible world model and the traditional probability distribution similarity in analyzing uncertain data.
Short Bio: Jian Pei is currently an Associate Professor and the Associate Director (Research and Industry Relations) at the School of Computing Science, Simon Fraser University, Canada. He is interested in developing effective and efficient data analysis techniques for novel data intensive applications, including data mining, Web search, information retrieval, data warehousing, online analytical processing, and database systems, as well as their applications in social networks, health-informatics, business intelligence and bioinformatics. His research has been well supported by numerous government funding agencies and industry partners. He has published prolifically and served actively in the research community. He is the recipients of several prestigious awards and honors.