A Comparative Analysis of Computational Complexity in Genetic Algorithm Variants: Evaluating Time and Space Trade-offs

Inspired by the evolution principle, GA becomes a more powerful optimization technique, although the authors have already implemented them in a rather different way in other algorithms; hence their differences concerning computational efficiency are extremely visible (Goldberg, 1989) [1]. This work aims to evaluate the computational complexity of the three GA variants: the Standard Genetic Algorithm (SGA), the Elitist Genetic Algorithm (EGA), and the Parallel Genetic Algorithm (PGA) (Mitchell, 1998) [2]. It will study computational complexity in terms of time (with respect to speed of convergence), number of function evaluations, and space complexity in terms of memory size, and memory needed to store the population (Deb, 2001) [3]. The experiment results show how time and memory consumption are related in executing it, hence providing insight on selecting the most appropriate type of GA when the resources used by environments are of paramount importance (Eiben & Smith, 2015) [6]. This might be followed by work on hybrid strategies and adaptive approaches that enhance performance further in GA algorithms (Srinivas & Patnaik, 1994) [12].