Numerical Optimization and Data Science (Master in Finance, LM in Mathematics) a.a. 2023 - 24
Start date: March 5, 2024
Tuesday 16 - 18, room B2
Thursday 16 - 18, room B4
Texts for further information:
J. Nocedal, S. J. Wright, Numerical Optimization (2nd edn.). Springer, 2006
M. J. Kochenderfer, T. A. Wheeler, Algorithms for Optimization, MIT Press, 2019
Ott1: Derivative-free methods Optimization test functions (wiki)
(Ott5: nonlinear Conjugate Gradient methods) skipped this year
Ott8: Applications to Artificial Neural Networks and Deep Learning
Matlab sessions:
Lab2: Newton's method (possibile code: main driver, calling the Newton function)
Lab3 Line Search methods using this backtracking version with Armijo rule
(tip: write a matlab function [xk1,alphak]= backtrack(fun,xk,gk,dk) and use the backtracking parameters sigma=1.e-4; rho=1/4; alphamin=1.e-5; alpha0 = 1 )
Lab4: Quasi-Newton (some solutions: out_QN_m1m1 out_QN_22 out_QN_05m05)
(Lab5: Nonlinear CG) skipped this year
Lab6: study of all methods seen so far for the Rosenbrock function (some solutions: out_QN_m1m1_rosenbrock out_QN_0505_rosenbrock
Lab7: Trust Region for Rosenbrock function (some solutions)
Lab8: Gauss-Newton and Levenberg-Marquart methods for nonlinear Least Square problems
Lab9: ANN