Sparse Regularized Fuzzy Regression

Applicable Analysis and Discrete Mathematics

article

Danilo Rapaic¹  Lidija Krstanovic²  Nebojša Ralevic¹  Ratko Obradovic²  Djuro Klipa¹ 

[1] University of Novi Sad, Faculty of Technical Sciences;University of Novi Sad, Faculty of Technical Sciences, Department of Fundamentals Sciences, Chair of Engineering Animation

关键词: MAP estimate;    robust statistics;    Huber norm;    sparse regularization;    Fuzzy regression;   

DOI  :  10.2298/AADM171227021R

学科分类：社会科学、人文和艺术（综合）

来源: Univerzitet u Beogradu * Elektrotehnicki Fakultet / University of Belgrade, Faculty of Electrical Engineering

【 摘 要 】

In this work, we focus on two things: First, in addition to the data measurement uncertainty, we develop a novel probabilistic model by imposing theadditive noise in the classical fuzzy regression model. We obtain the baselineLS estimation as the maximum likelihood estimation for regression parameters. Moreover, by assuming the heavy tail distribution and by introducingthe Huber norm instead of square in the cost function, we obtain more general robust fuzzy M-estimator, much more suitable for modeling the outliersoften present in the data sets.

【 授权许可】

Unknown