A Hybrid Approach: Combining Genetic Algorithms and Machine Learningfor Function Optimization
Prashant Kumar · Journal of Computer Allied Intelligence (JCAI). · 2025
This study examines the use of an evolutionary method to enhance the Sphere benchmark function, an acknowledged continuous optimization challenge. The algorithm takes a genetic approach, using techniques including mutation, one-point crossover, and tournament selection. It also combines a machine learning element by developing a simple linear regression model that can be used to forecast fitness values determined by attributes of individuals. The study compares the results of two different iterations to examine the algorithm's performance throughout several runs. A population of 100 individuals with 10 traits each endures selection, crossover, and mutation over the course of 100 generations. The best values for fitness across generations for each run are shown in Matplotlib to show the algorithm's convergence behaviour. Results show that the algorithm works effectively in locating the best solutions to the Sphere benchmark function. The algorithm's framework, parameters, and convergent behaviour are all described in the abstract, which qualifies it for future study in adaptive algorithms and optimization approaches