Modeling of fuzzy integral based nonlinear multi-regressions systems with QPSO-GS

A nonlinear multi-regression based on fuzzy integral (NAFI) model that include outliers under inherent interaction among feature attributes is considered in this paper. The modeling of the proposed model is also performed via a modified algorithm base on particle swarm optimization with quantum behavior (MQPSO) and the high breakdown value estimator, least trimmed squares (LTS). That is, we successfully integrate mechanisms of the genetic algorithm and the simulated annealing into the QPSO algorithm to estimate parameters of the NAFI model; meanwhile, the LTS estimator is also introduced to filter out outliers. From simulation results, the proposed MQPSO algorithm with LTS estimator (named QPSO-GS) readily corrects the deviation caused by outliers and swiftly achieves convergences on estimating the parameters of the proposed NAFI model with outliers.