A Machine Learning based Approch to Classify Emotions from Music

Music has the profound ability to evoke a wide spectrum of emotions, from joy and excitement to sadness and melancholy. A vital field of study with uses in advertising, film soundtracks, music therapy, and emotional analysis, emotion classification in music is the identification and classification of these emotional reactions. This work investigates how machine learning methods might precisely categorize musical emotions into happy, sad, calm, and energetic groups. Key audio features including Mel-Frequency Cepstral Coefficients (MFCCs), spectral contrast, and rhythm-based characteristics are the main emphasis of the feature extracting process. We apply Random Forest (RF), K-Nearest Neighbors (KNN), and Support Vector Machines (SVM) among other supervised machine learning techniques to assess classification performance. Cross-valuation guarantees model dependability; performance is measured with accuracy, F1-score, recall, and precision. Experimental findings show that the suggested method provides a strong basis for emotional-based music analysis since it efficiently labels music files with great accuracy.