Analytic Continued Fractions for Regression: A Memetic Algorithm Approach. (1st October 2021)

• Record Type: Journal Article
• Authors: Moscato, Pablo; Sun, Haoyuan; Haque, Mohammad Nazmul
• Is Part Of: Expert systems with applications. Volume 179(2021)
• Publication Date: 2021-10-01
• Subjects: Symbolic regression -- Memetic algorithm -- Analytic continued fractions; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2021.115018
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: A regression method employing analytic continued fractions as novel representation. CFR ranked 1st among the 16 methods for generalisation performances on 94 datasets. Statistically comparable to the best 3 algorithms (eplex-1m, xgboost & grad-boost). CFR results have been obtained with a limited level of parameter tuning. Abstract: We present an approach for regression problems that employs analytic continued fractions as a novel representation. Comparative computational results using a memetic algorithm are reported in this work. Our experiments included fifteen other different machine learning approaches including five genetic programming methods for symbolic regression and ten machine learning methods. The comparison on training and test generalization was performed using 94 datasets of the Penn State Machine Learning Benchmark. The statistical tests showed that the generalization results using analytic continued fractions provide a powerful and interesting new alternative in the quest for compact and interpretable mathematical models for artificial intelligence.