Abstract
Classification, clustering, and frequent pattern mining are the most common tasks in data mining. Almost all of these tasks can be completed using machine learning algorithms, but sometimes even these algorithms will not perform better. Deep learning is currently used in solving these data mining problems which very high accuracy. However, a recent report from the Massachusetts Institute of Technology states that computational limits of deep learning have reached. The report also stated that improvements can be achieved by using some optimization framework. Metaheuristic algorithms provide a framework for solving optimization problems and it is shown in this chapter some times metaheuristic algorithms outperform these machine learning algorithms and with less complexity. This chapter discusses different methods by which some common metaheuristic algorithms like Ant colony optimization, Genetic Algorithms, and Particle Swarm Optimization can be used to perform the different data mining tasks effectively.
