This book provides a comprehensive exploration of population-based algorithms through a machine learning lens, focusing on evolutionary and swarm intelligence methods. While the book presents the core algorithms in the field, it also is written to cover topics relevant to difficult optimization problems, such as optimization under uncertainty, optimization in high dimensional spaces, and optimization in dynamic environments. Readers will gain insights into these algorithms by considering issues such as the roles and impacts of representation and preference bias in search, issues around hyperparameter tuning and optimization, design of hybrid methods that incorporate more traditional methods, and population-based methods for solving reinforcement learning (i.e., control and sequential decision making) problems. The book is also written from the perspective of guiding a young investigator who is looking to perform research in the field of evolutionary and swarm-based algorithms. Thus, the book serves an essential resource for graduate students, early-career researchers, and academic professionals in computer science, artificial intelligence, and related fields. Approaches population-based algorithms from the perspective of their relationships to methods in machine learning. Discusses evolutionary algorithms, including genetic algorithms, genetic programming, and differential evolution, with a focus on issues in single and multi-objective optimization. Covers swarm-based algorithms such as ant colony optimization and particle swarm optimization, along with alternative swarm intelligence methods. Explains theoretical foundations, including convergence properties and computational complexity of population-based algorithms. Introduces advanced topics like co-evolution, combinatorial optimization, self-adaptive algorithms, and hybrid/memetic approaches. Explores applications in reinforcement learning, optimization under uncertainty, and search in dynamic environments.
