Feature Selection Using Improved Forest Optimization Algorithm

Feature selection is a very popular topic in the field of data mining and machine learning. In 2016, the feature selection using forest optimization algorithm (FSFOA) was proposed, which had a better classification performance and dimensionality reduction ability. However, there are some shortcomings in FSFOA. In this article, Feature Selection using Improved Forest Optimization Algorithm (FSIFOA) is proposed, which aims at solving the problems of FSFOA during the stages of random initialization, forming the candidate population and updating the best tree. The proposed FSIFOA is compared with some other methods including FSFOA, NSM, PSO and other algorithms. The experimental results show that FSIFOA can improve the classification accuracy of classifiers in medium and large dimension datasets. Also, the dimensionality reduction of the FSIFOA is compared with other comparable methods.