BCAM is the research center on applied mathematics created with the support of the Basque Government and the University of the Basque Country, which aims to strengthen the Basque science and technology system, by performing interdisciplinary research in the frontiers of mathematics, talented scientists’ training and attraction, so the excellence of our results are recognized by the Society.
The team involved in the JRL on behalf of BCAM is the Data Science group, led by Prof. Jose A. Lozano. The objective in this line is to design new methods for learning probabilistic graphical models in general, and Bayesian networks in particular. These methods will be a hybrid between classical operations research techniques and metaheuristic methods. We will apply the developed techniques in the solution of real data analysis tasks.
This area of research combines many mathematical methodologies. First of all are probability theory and statistics. The final objective is to learn a probability distribution. It also deals with graph theory as probabilistic graphical models are represented using graphs. A third topic used in the area is optimization. Most of the processes are based in optimizing a score, and therefore the area of combinatorial optimization, in particular is basic for the topic.