

소개
This work introduces a system for the automatic classification of music mood em- ploying the Random Forest algorithm and audio feature extraction with the Librosa library. The suggested solution utilizes prominent audio descriptors, Mel-frequency cep- stral coefficients (MFCCs), and Mel spectrograms, in order to achieve the timbral and spectral qualities of music pieces. These audio features are obtained from a cleaned dataset of tagged audio files reflecting different mood labels.



