SOLVING FEATURE SELECTION PROBLEM USING BACTERIAL FORAGING ALGORITHM

Abstract

Feature selection has become important and beneficial when the amount of data and information is large. Sometimes, reduction of features can improve the quality of prediction and even be a necessary embedded step of the prediction algorithm. Further improvements in feature selection will affect a wide array of applications in fields such as biomedical, pattern recognition, machine learning, or signal processing. Bacterial Foraging Algorithm (BFA) is one of the powerful bio-inspired optimization algorithms, which attempt to imitate the single and groups of E. Coli bacteria. In BFA algorithm, sets of bacteria try to forage towards a nutrient rich medium to get more nutrients. In this scheme, an objective function is posed as the effort or a cost incurred by the bacteria in search of food. BFA with a Naive Bayes classifier has been presented in this work. The performance of the proposed feature selection algorithm was investigated using six benchmark datasets from different domains and was compared to three other well-known feature selection algorithms. Discussion focused on two perspectives: number of features and classification accuracy. The results showed that proposed work outperformed other algorithms in selecting lower number of features, hence removing irrelevant, redundant, or noisy features while maintaining the classification accuracy.