This study presents a classification model that incorporates significant enhancements based on the Bayesian method and genetic algorithm (BGA). Firstly, the prior probabilities in each iteration are determined using the ratio of the number of elements in each group, obtained through clustering techniques, to the total number of elements in the training set. Secondly, an automatic selection process optimizes the training set to minimize classification errors. Finally, the traditional genetic algorithm operators are improved by utilizing the Bayes error as the objective function. These improvements combine to create an effective classification model. Additionally, the BGA demonstrates effective performance on real data using the established MATLAB procedure. A numerical example illustrates the superiority of the proposed algorithm compared to existing methods. The study also applies the BGA for image classification using the Gabor filter, which extracts essential image features. The proposed model outperforms popular methods in classifying various numerical and image datasets. These applications highlight the potential of this study in real-world scenarios.