Lecture notes.
Book Metodi di Ottimizzazione non vincolata, Grippo-Sciandrone, Springer 2011
Learning Objectives
1) Ability to formulate nonlinear optimization problems.
2) Knowledge of algorithms for large-scale optimization, sparse optimization, newtork equilibrium, machine learning,
multi-agent optimization, multiobjective optimization.
3) Ability to use e adapt standard nonlinear optimization algorithms to specific practical contexts.
Prerequisites
Linear algebra and analysis
Teaching Methods
Frontal lectures
Type of Assessment
Written (or oral) examination having theoretical questions:
- to verify the ability to formulate optimization problems;
- to verify the ability to use and adapt standard nonlinear optimization algorithms;
- to verify the knowledge of algorithms for classes of complex optimization problems.
Course program
Decomposition algorithms for unconstrained and constrained optimization. Decomposition algorithms for machine learning. Steepest descent methods for multiobjective optimization. Nash equilibrium: models and algorithms. . LASSO methods for sparse optimization. Concave programming for minimizing the zero norm. Algorithms for multiobjective optimization.