Course description:
Participants learn basic concepts, their theoretical foundation, and the most common algorithms used in machine learning and artificial intelligence. After completing the module, participants should understand strengths and limitations of the different paradigms, should be able to correctly and successfully apply methods and algorithms to real world problems, should be aware of performance criteria, and should be able to critically evaluate results obtained with those methods. Participants should also be able to modify algorithms to new tasks at hand as well as to develop new algorithms according to the paradigms presented in this course.
Contents include:
Artificial neural networks: Connectionist neurons, the multilayer perceptron, radial basis function networks, learning by empirical risk minimization, gradient-based optimization, overfitting and underfitting. Learning theories and support vector machines: statistical learning, learning by structural risk minimization.
Probabilistic methods: Bayesian inference and neural networks, generative models. Projections methods. Principal Component Analysis, Independent Component Analysis and blind source separation. Stochastic optimization. Clustering and embedding.