CONTINUOUS TIME BAYESIAN NETWORKS FOR MINING STREAMING DATA

Abstract: Streaming data are relevant in finance, computer science, and engineering while they are becoming increasingly important in medicine and biology. In particular, classification, clustering and structural learning of streaming data are receiving increasing attention. These tasks require algorithms and models capable to represent dynamic, sequence and time. Dynamic Bayesian networks and hidden Markov models are used to analyze streaming data. However, these models are concerned with equally spaced time data and thus suffer from several limitations because it is not clear how to discretize timestamps. The talk introduces continuous time Bayesian networks and continuous time Bayesian networks classifiers. Algorithms for parametric and structural learning of continuous time Bayesian network models to solve classification, clustering and structural learning based on multivariate discrete state continuous time trajectories are described. Stationary and non-stationary continuous time Bayesian networks are presented together with their structural learning based on the marginal likelihood approach. Numerical experiments concerning real world applications in finance, biomedicine, neurology and biology are presented.

CV: Prof Fabio Stella is an associate professor at the Dipartimento di Informatica, Sistemistica e Comunicazione of the Università degli Studi di Milano-Bicocca. His research focuses on models and algorithms for data analysis and decision making under uncertainty in the areas of Business Intelligence, Data and Text Mining and Computational Finance. In the winter term 2014/15 he is giving the course 625.605 – Business Intelligence in Klagenfurt.

image001

Posted in TEWI-Kolloquium | Kommentare deaktiviert für CONTINUOUS TIME BAYESIAN NETWORKS FOR MINING STREAMING DATA
RSS
EMAIL