Machine Learning is a branch of Artificial Intelligence that tries to improve the behaviour of an agent using the input supervision. This supervision can be in forms of the labeled data or reinforcement  signals from the environment. In this group we focus on the methods that use the labeled data as the supervision to solve the Pattern Recognition problem.

Our main research activities in this field consist of Semi-Supervised Learning methods, Meta Learnings such as Kernel Learning and Neighborhood Graph Construction, Kernel based Learning methods such as SVM and Gaussian Processes.