Abstract:
Mixture distributions in general are an important tool in complex data modelling. One of the applications is in target tracking systems, where the state and measurement model are represented by mixture distributions. The update step (Bayesian inference) of the tracking procedure yields, through multiplication of mixtures, an exponential growth in the number of components of the mixture representing the state. Therefore, a component reduction procedure is a mandatory part of every feasible tracking system. In this talk we will present a systematic information theoretic approach for component reduction in mixtures of the exponential family, with emphasis on the von Mises mixtures.
CV:
Ivan Marković received the B.Sc. degree in Electrical Engineering from the Faculty of Electrical Engineering and Computing (FER Zagreb), University of Zagreb, Croatia in 2008. He is currently employed at the FER, Zagreb, as a Research Assistant and a full-time Ph.D. student funded by the Ministry of Science, Education and Sport, Republic of Croatia. His main research interests are in mobile robotics, estimation theory and human–robot interaction. He is also a Teaching Assistant at several undergraduate courses. During his undergraduate studies, for outstanding academic achievements, he received the ‘‘Institute for Nuclear Technology Award’’ and the ‘‘Josip Lončar Award’’ in 2007 and 2006 respectively. He also serves as a technical editor of the Automatika journal. In 2001/2002 he spent his junior high school year in the USA as a scholar funded by the American Secondary Schools for International Students and Teachers organisation.
Mario Bukal is research associate at the Research Centre for Advanced Cooperative Systems (ACROSS), Faculty of Electrical Engineering and Computing (FER), University of Zagreb. He received the M.Sc. degree in Mathematics from the Department of Mathematics, Faculty of Science (PMF), University of Zagreb in 2008, and the PhD degree in Applied Mathematics from the Faculty of Mathematics and Geoinformation, Vienna University of Technology, Austria in 2012. His research interests include nonlinear partial differential equations, dynamical systems, information theory and optimization problems.