Klasifikasi Mood pada Musik Pop dan Jazz dengan Menggunakan Mel Frequency Cepstral Coefficients dan K-Nearest Neighbor

Abstract

This research discusses mood classification in pop and jazz music using Mel Frequency Cepstral Coefficients (MFCC) and the K-Nearest Neighbor (KNN) algorithm. The dataset used consists of 900 songs with mood labels angry, happy, relaxed, and sad obtained from Kaggle. The data was processed by extracting 13 MFCC features and then continuing with classification using KNN. The research results show that the best accuracy reaches 64% with K=9. Accuracy at K=7 obtained a value of 60%, while at K=11 an accuracy of 58% was obtained. Evaluation was carried out using accuracy, precision, recall and f1-score metrics, with the best results found at K=9. This research emphasizes the importance of selecting K parameters for optimizing mood classification models.

Keywords: Mood Clasification, MFCC, K-Nearest Neighor, Music Emotion Recognition