Course type: compulsory
Lecturer: Assistant Professor Erik Štrumbelj
 

For the timeline see Curriculum.
 

Content (Syllabus outline):
•    Bayesian methods: Gaussian processes, Dirichlet processes, MCMC methods, variational inference.
•    Deep learning: Boltzmann machines, Autoencoders, Convolutional neural networks.
 
•    Computational learning theory: PAC learning, VC dimension.
•    Other select topics: multi-kernel learning, multi-task learning, reinforcement learning.

Objectives and competences:

The main objective is to familliarize the students with advanced machine learning methods. Practical applications and the mathematical and algorithmic background are equally important.
 

Intended learning outcomes:

Knowledge and understanding: Understanding advanced machine learning methods and the underlying mathematics and algorithms.
Application: Advanced machine learning methods can be applied to solve the most demanding practical problems in data analysis. The concepts covered by this course are also fundamental to methodological and theoretical research in machine learning.
Reflection: understanding of the theory on the basis of examples of application.
Transferable skills: Analytical and research ability. Ability to solve practical data analysis problems.