Vue d'ensemble
G茅nie (CCE) : Systems modeling and optimization: Basic modeling concepts, parametrization, data fit, design space sampling, mathematical problem formulation (linear and nonlinear), monotonicity and boundedness, constraint activity. Unconstrained optimization problems: Function approximation, optimality conditions, convexity, line search, gradient-based methods, Newton method. Constrained optimization problems: Feasibility, Lagrange multipliers, generalized reduced gradient method, Karush-KuhnTucker (KKT) conditions. Algorithms and software: Global and local convergence terminology and properties, termination criteria, scaling, commercial tools (Matlab optimization toolbox), derivative-free algorithms (search methods and genetic algorithms).
Trimestres : 脡t茅 2016
Charg茅s de cours : Kokkolaras, Michael (Summer)
15 hours of classroom instruction, 9 hours of additional work (project). Some basic understanding of calculus and linear algebra as well as programming and computer skills (e.g. Matlab) is helpful, but is not required as the course module will introduce and review all necessary concepts.