This chapter introduces Genetic Algorithms (GA) as a powerful evolutionary optimization technique for neural network training, particularly addressing limitations of traditional backpropagation methods like local optima convergence. It covers fundamental GA concepts including selection, crossover, and mutation operators, and their application in optimizing neural network weights and biases. The chapter demonstrates practical implementation through a comprehensive Python example solving the XOR classification problem, combining BP neural networks with genetic optimization. The implementation includes fitness evaluation, population evolution, and performance visualization. Analysis reveals steady fitness improvement over generations and discusses trade-offs between precision and recall in classification performance. The chapter concludes with practical recommendations for parameter tuning and architectural improvements to enhance GA-based neural network optimization.