15 February 2018
8 hours on 2 days (15/02 - 16/02)
Louvain-la-Neuve
ISBA - C115 (Seminar Room Bernoulli) and C045 (Salle Gauss)
Statistics and optimisation by Professor Sylvain Sardy, Université de Genève
Dates
15/02/2018 and 16/02/2018 at ISBA, UCL, Louvain-la-Neuve
Registration
Please register on-line : Registration Form
Course Title and Content
Statistics and optimisation (Computational aspects of regularisation methods for high-dimensional data problems)
Hour 1: Motivation for doing optimization in Statistics with numerous examples
Hour 2: Formalization of the optimization problem, optimiality conditions, existence and uniqueness theorems
Hour 3: Univariate optimization and general iterative algorithms formultivariate optimisation (Steepest descent, Newton, relaxation)
Hour 4-5: Convex sets and functions, sub gradient, KKT conditions
Hours 6-8: Solving optimization problems of estimators including ridge regression, waveshrink, basis pursuit, robust waveshrink, compressed sensing, total variation smoothing, segmentation and density estimation (fused lasso) and more.
Hour 2: Formalization of the optimization problem, optimiality conditions, existence and uniqueness theorems
Hour 3: Univariate optimization and general iterative algorithms formultivariate optimisation (Steepest descent, Newton, relaxation)
Hour 4-5: Convex sets and functions, sub gradient, KKT conditions
Hours 6-8: Solving optimization problems of estimators including ridge regression, waveshrink, basis pursuit, robust waveshrink, compressed sensing, total variation smoothing, segmentation and density estimation (fused lasso) and more.
Venue and schedule
Thursday 15/02/2018
9h30 - 12h30 : C 115
14h00 - 16h00 : C115
9h30 - 12h30 : C 115
14h00 - 16h00 : C115
Friday 16/02/2018
9h00 - 12h00 : C045
9h00 - 12h00 : C045
Categories Events: