Neural Networks – A.A. 2012 / 2013
Fundamentals of Metaheuristics
A three-part seminar on metaheuristics, i.e., general optimization algorithms with few assumptions on the function being optimized and no guarantee on global convergence. Two very simple examples in Matlab are provided for each part:
For more advanced material, refer to the following ebook, from which part of the seminar was built: Essentials of Metaheuristics. Another good textbook is Global Optimization Algorithms: Theory and Applications.
Statistical Learning Theory
A small introduction to Statistical Learning Theory, following the classical exposition from Vapnik.
Differents projects are avalaible for the final exam, including: Multi-Objective Optimization, Coevolutionary Methods, Island Models, Semi-Supervised algorithms, sparsity-aware methods, Deep structures, Kernel Adaptive Filtering and others.