A Modified Ant Lion Optimization Method and Its Application for Instance Reduction Problem in Balanced and Imbalanced Data

Instance reduction is a pre-processing step devised to improve the task of classification. algorithms search for reduced set instances mitigate low computational efficiency and high storage requirements. Hence, finding optimal subset utmost importance. Metaheuristic techniques are used as potential application. Antlion optimization (ALO) recent metaheuristic algorithm that simulates antlion’s foraging performance in attacking ants. However, ALO suffers from local optima stagnation slow convergence speed some problems. In this study, new modified antlion (MALO) recommended primary by adding parameter depends on length each ant while revising position. Furthermore, suggested MALO adapted challenge instance obtain better results terms many metrics. The based twenty-three benchmark functions at 500 iterations thirteen 1000 demonstrate proposed escapes provides rate compared basic well-known algorithms. addition, 15 balanced imbalanced datasets 18 oversampled show method can statistically outperform has strong competitiveness against other comparative four measures: Accuracy, Balanced Accuracy (BACC), Geometric mean (G-mean), Area Under Curve (AUC) addition run time. increment BACC, G-mean, AUC rates up 7%, 3%, 15%, 9%, respectively, over keeping less