Foundations of AI and Optimisation Methods

The course comprises two parts. The first part, which constitutes approximately 60% of the course, treats basic concepts and paradigms of artificial intelligence and machine learning, for instance hypothesis space, generalisation error, and limitations. This first part of the course also includes practical components, such as design of algorithms for linear and logistic regression and support vector machines, and practical aspects of machine learning, for instance normalisation and cross-validation. The second part of the course treats optimisation with and without constraints. Among other things, this includes showing how stochastic gradient descent can be used for optimisation without constraints.