Discriminating and Clustering Ordered Permutations Using Neural Network and Potential Applications in Neural Network-Guided Metaheuristics

Adaptive Resonance Theory (ART) neural network has been used in many applications due to its fastadaptable learning process and stable operations. In this work, we present a technique for discriminating and clustering ordered permutation using ART-1 and Improved-ART-1. In the process, we developed a novel technique for converting ordered permutations to binary vectors to cluster them using ART. The performances of ART-1 and Improved-ART-1 have been investigated, and the proposed binary conversion methods were evaluated under varying parameters and problem sizes. Three performance indicators, i.e., misclassification, duster homogeneity, and average distance are considered in the analysis. The numerical results indicate the superiority of one of the proposed binary conversion techniques over the other and Improved-ART-1 over ART-1. Moreover, potential applications of the proposed technique in developing ANN guided metaheuristics to solve problems whose solutions are ordered permutations are discussed. A case study in solving flexible flow shop scheduling using ANN guided Genetic Algorithm is also presented.