Robust Fuzzy Clustering Algorithms for Change-Point Regression Models

• DOI: 10.1142/s0218488520500300
• Publication Date: 2020-09-03T11:21:10Z
• Authors: Kang-Ping Lu; Shao-Tung Chang
• Keywords: Robust regression; Leverage (statistics); Robust Statistics; Robustness

This article presents a robust fuzzy procedure for estimating change-point regression models. We propose incorporating the algorithm with M-estimation technique estimations. The c partitions concept is embedded into model so c-regressions and c-means clustering can be employed to obtain estimates of change-points parameters. M estimation criterion used make estimators presence outliers heavy-tailed distributions. create two algorithms named FCH FCT by using Huber’s Tukey’s functions as respectively. Extensive experiments numerical real examples are provided demonstrating effectiveness superiority proposed algorithms. experimental results show resistant atypical observations outperform existing methods. generally comparable but performs better in extremely high leverage Real data applications practical usefulness method.