A0860
Title: Robust estimation of mixed models for energy production
Authors: Luis Angel Garcia-Escudero - Universidad de Valladolid (Spain)
Fabrizio Laurini - University of Parma (Italy) [presenting]
Agustin Mayo-Iscar - Universidad de Valladolid (Spain)
Abstract: Parameter estimation for mixed models can be severely biased by outliers. The outliers can be in the conditional distribution of the response variable or even in the explanatory variables. The proposal is to mitigate the effect of outliers by introducing robust estimations by trimming a fixed proportion of observations. The level of trimming is an input hyperparameter, and it can be adjusted and monitored with proper diagnostics. Some results are shown with artificial data, and it is compared with available competitors. The proposed robust method is also applied to real data, based on electricity production data in Europe.
