Feature selection software development using Artificial Bee Colony on DNA microarray data

Abstract

DNA Microarray data is a high-dimensional data that enables the researchers to analyze the expression of many genes in a single reaction quickly and in an efficient manner. Its characteristics such as small sample size, class imbalance, and data complexity causes it difficult to classified. Feature selection is a process that automatically selects features that are most relevant to the predictive modeling in dataset. This research aims at investigating, implementing, and analyzing a feature selection method using the Artificial Bee Colony (ABC) approach. The result is compared with other evolution algorithms, which is Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). The result is that feature selection using ABC has a better result at classification using k-Nearest Neighbor (k-NN) and Decision Tree (DT), but has a slightly higher fracture of features compared to GA and PSO algorithms.