Optimizing the regularization parameters in sparse modelling

VICTORIA PETERSON; RUBÉN SPIES

Vancouver

Congreso; LXAI Workshop- NeurIPS 2019; 2019

Tikhonov functionals are commonly used as regularization strategies for severelyill-posed inverse problems. Besides the type of penalization induced into the solution, the proper selection of the regularization parameters is of utmost importancefor accurate solutions. In this work, we analyze several data-driven regularization parameters estimation methods in a mixed-term discriminative framework.Numerical results for P300 detection in Brain-Computer Interfaces classificationare presented, showing the impact of regularization parameter estimation intoclassification performance.