Music Style Recognition using Mel-Frequency Cepstral Coefficients and Principal Component Analysis Enhanced Features with Support Vector Machine

Music style recognition is a key application in audio classification, recommendation systems, and music information retrieval. Most conventional approaches depend on handcrafted feature extraction or neural networks with the need for extensive training and large datasets. Nevertheless, current systems have low accuracy when handling overlapping genres or noisy audio signals. This study puts forward the hybrid classification system that integrates Support Vector Machine (SVM) with Mel-Frequency Cepstral Coefficients (MFCC) feature extraction and Principal Component Analysis (PCA) for reducing dimension, which is more effective and efficient. SVM is utilized because it is strong in dealing with small and high-dimensional data sets. The proposed system is compared with standard classifiers KNN and Decision Trees. Results demonstrate the new system attains 97.4% accuracy and an F1-score of 96.8%, which is significantly higher than conventional models. The improved performance can be credited to the discriminative nature of MFCC features and PCA's capacity to remove redundant data, allowing the SVM classifier to have better distinction between genres. This hybrid model exhibits enhanced generalization, accuracy, and reliability in music style classification tasks.