Feature Selection of Microarray Data using Bacterial Foraging Optimization

摘要

This study aims at finding the genes responsible for lung cancer using bacterial foraging optimization. Microarray datasets can be used to predict the presence of cancer, the type of cancer, its stage etc. Microarray datasets available have large number of features and few samples. Bacterial foraging optimization algorithm has been used in our study for feature selection on lung cancer dataset. BFO algorithm selects few genes from the available set. The reduced dataset is then used for designing a classifier using support vector machines which gives an accuracy of about 99% classifying only one sample inaccurately.