Music genre classification is an important multimedia research domain, including aspects of music piece representation, distances between genres, and categorization of music databases. The objective of this study was to develop a model for automatic classification of musical genres from audio data by using features from low-level time and frequency domains. These features can highlight the differences between different genres. In the model, feature selection is performed using a genetic algorithm (GA), and the resulting dataset is classified using the k-nearest neighbor (KNN), naive Bayes classifier (NBC), and support vector machine (SVM) learning methods. Tenfold cross-validation is used to obtain the optimal f-measure value. In this study, the data were obtained from the GTZAN genre collection datasets. In the performance evaluation, it was found that the GA-based feature selection strategy can improve the F-measure rate from 5% to 20% for the KNN, NBC, and SVM-based algorithms. In addition, the proposed SVM-GA algorithm can exactly better than other comparison algorithms.