Development of Models and Methods for Incorporating Knowledge to Support Vector Machines

The first goal of the project is to develop machine learning models with uncertain knowledge extended by mathematical structures and combined with theoretical models in the form of differential equations. The second goal is to develop methods of incorporating prior knowledge to support vector machines (SVM) based on proposed models.

The main hypotheses of this project are: 1) incorporating knowledge about mathematical structures and differential equations to machine learning models will allow us to improve generalization bounds 2) there exist efficient machine learning methods with incorporated prior knowledge about mathematical structures and differential equations for solving real world problems.