Swarm Intelligence-based Decision Trees Induction for Classification — A Brief Analysis

Abstract

Decision trees are popular machine learning classifiers, they accurately represent the data in a simple manner that closely resembles human reasoning. Since inducing the optimal decision tree is a NP-hard problem, numerous traditional heuristic-based approaches were introduced to tackle it. However, due to the present data explosion, these greedy local methods did not guarantee the induction of an optimal tree. To address this issue, swarm intelligence algorithms have been currently applied to navigate the search space more appropriately, seeking optimal decision trees. The aim of this research study is to give an analysis overview of the most up-to-date existing swarm-based decision trees induction techniques in a shape of a comparative study, where we discuss the different basics, features, characteristics and results. This survey will serve as a guide for the researches community. However, due to the present data explosion, these greedy local methods did not guarantee the induction of an optimal tree. To address this issue, swarm intelligence algorithms have been currently applied to navigate the search space more appropriately, seeking optimal decision trees. The aim of this research study is to give an analysis overview of the most up-to-date existing swarm-based decision trees induction techniques in a shape of a comparative study, where we discuss the different basics, features, characteristics and results. This survey will serve as a guide for the researches community. The aim of this research study is to give an analysis overview of the most up-to-date existing swarm-based decision trees induction techniques in a shape of a comparative study, where we discuss the different basics, features, characteristics and results. This survey will serve as a guide for the researches community.