Chapter 5 - Whale optimization algorithm and its application in machine learning

Abstract

Machine learning is the most well-regarded and fast-growing field in artificial intelligence. Machine learning methods have been successfully applied to solve many real-world problems, such as producing recommendations, image recognition, sentiment analysis, fraud detection, and so on. However, to make machine learning algorithms efficient, some challenges should be handled. For instance, the efficiency of machine learning algorithms highly depends on finding optimal learning parameters and assigning optimal values to the hyperparameters of machine learning methods. To tackle these challenges, we can utilize optimization algorithms. Among many types of optimization algorithms, nature-inspired algorithms are very promising methods because they are simple to implement, efficient for global search, and are unable to escape from local and fastest-optima. This chapter investigates how the Whale optimization algorithm, one of the nature-inspired algorithms, can be applied to various machine learning problems such as feature selection, tuning the parameters of SVM classifiers and K-means clustering, and tuning the hyperparameters of artificial neural networks to improve their performance.