A Robust Multi-Objective Feature Selection Model Based on Local Neighborhood Multi-Verse Optimization

Classification tasks often include, among the large number of features to be processed in datasets, many irrelevant and redundant ones, which can even decrease efficiency classifiers. Feature Selection (FS) is most common preprocessing technique utilized overcome drawbacks high dimensionality datasets has two conflicting objectives: The first function aims maximize classification performance or reduce error rate classifier. In contrast, second designed minimize features. However, majority wrapper FS techniques are developed for single-objective scenarios. Multi-verse optimizer (MVO) considered as one well-regarded optimization approaches recent years. this paper, binary multi-objective variant MVO (MOMVO) proposed deal with feature selection tasks. standard MOMVO suffers from local optima stagnation, so we propose an improved issue using memory concept personal best universes. experimental results comparisons indicate that approach effectively eliminate and/or maintain a minimum when dealing different compared popular techniques. Furthermore, 14 benchmark showed outperforms stat-of-art algorithms selection.