期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2020
卷号：98
期号：23
页码：3823-3833
出版社：Journal of Theoretical and Applied
摘要：In this paper, two types of controller for a mobile robot with the wall-following task are statistically compared. One of them is a Multi-Layer Perceptron (MLP) while the other one is an Adaptive-Network-based Fuzzy Inference System (ANFIS). Here, such controllers are named: EL-MLP and EL-ANFIS, because they were trained by means of an analytical method known as Extreme Learning Machine (ELM). They were structurally optimized with a statistical method known as Analysis of Variance (ANOVA), and a t-Test between two populations with the best exemplars of each type of controller, demonstrates statistically that EL-ANFIS generalizes better than EL-MLP, due to its validation error mean, and variance, are significantly lower.
关键词：Multi-Layer Perceptron (MLP);Adaptive-Network-based Fuzzy Inference System (ANFIS);Extreme Learning Machine (ELM);Analysis of Variance (ANOVA);Hypothesis test between two populations (t-Test);PD Control.
