Insights into the feature selection problem using local optima networks

Citation

Mostert W, Malan KM, Ochoa G & Engelbrecht AP (2019) Insights into the feature selection problem using local optima networks. In: Liefooghe A & Paquete L (eds.) Evolutionary Computation in Combinatorial Optimization. Lecture Notes in Computer Science, 11452. 19th European Conference on Evolutionary Computation in Combinatorial Optimization, EvoCOP 2019, held as part of EvoStar 2019, Leipzig, Germany, 24.04.2019-26.04.2019. Cham, Switzerland: Springer Verlag, pp. 147-162. https://doi.org/10.1007/978-3-030-16711-0_10

Abstract
The binary feature selection problem is investigated in this paper. Feature selection fitness landscape analysis is done, which allows for a better understanding of the behaviour of feature selection algorithms. Local optima networks are employed as a tool to visualise and characterise the fitness landscapes of the feature selection problem in the context of classification. An analysis of the fitness landscape global structure is provided, based on seven real-world datasets with up to 17 features. Formation of neutral global optima plateaus are shown to indicate the existence of irrelevant features in the datasets. Removal of irrelevant features resulted in a reduction of neutrality and the ratio of local optima to the size of the search space, resulting in improved performance of genetic algorithm search in finding the global optimum.

Keywords
Local optima networks; Feature selection; Fitness landscape analysis

People

Professor Gabriela Ochoa

Professor, Computing Science