Hybrid CNN and Logistic Regression Approach for Folk Song Classification

The summary of this research presents a thorough analysis of the performance of the classification model across five unique categories of folk songs: Ainam, Bianam, Bihunam, Gualporia Lokageet, and Kamrupia Lokageet. The precision, recall, and F1-score metrics, as detailed in Table 1, demonstrate strong performance with values ranging from 93.75% to 94.17% for all categories. The model, which was trained using 10,090 images, obtained an accuracy of 94.05% overall. These metrics indicate the model’s capability to accurately classify instances of each type of folk song. provides further insights into the dataset, including support counts, proportion within the dataset, and overall accuracy, all consistently recorded at 0.98 for all categories. This consistency highlights the reliability and robustness of the dataset in training and evaluating the classification model. This results indicate that the model achieved high precision in correctly identifying instances within each category, while maintaining consistent recall rates and F1-scores. These outcomes showcase the effectiveness of the methodology in handling various types of folk songs and offer valuable insights into the performance characteristics of the model. In conclusion, this study makes a significant contribution to the field of music classification by presenting a model that achieves high accuracy and reliability in distinguishing between different categories of folk songs. The metrics presented can serve as a reference point for future research in music classification, emphasizing the model’s potential for practical applications in music analysis and recommendation systems.