CBTS-SGL Webinar - Cool Things That One Can Do With Graphical Probabilistic Models - Dr. Marek Drudzel

Abstract

On January, 2021 the CBTS COE and the Stochastic Geomechanics Laboratory (SGL) organized a webinar with Dr. Marek Drudzel as a guest speaker to give a presentation on graphical probabilistic models, and their potential applications to real-world problems. The talk reviewed briefly the theoretical foundations of Bayesian networks and their applications to practical problems. Dr. Drudzel showed several flavors of Bayesian networks, such as discrete, continuous, and hybrid networks, as well as qualitative, and dynamic networks. He also reviewed some applications of Bayesian network models in diagnosis, prognosis, data analysis, and strategic planning. Part of the talk was based on live demonstration of the concepts using GeNIe, a software developed originally in Dr. Druzdzel's lab at the University of Pittsburgh.