Research on Multi-Feature Fusion Music Emotion Classification Method Under Cognitive Psychology

Music is one of the most essential forms of human communication, capable of conveying a wealth of emotional information. This paper proposes a novel multi-feature fusion algorithm for music emotion classification, combining features such as timbre, dynamics, melody, rhythm, pitch, range and time domain to capture the complexity of emotional expression in music. Grounded in cognitive psychology, the proposed method employs a Backpropagation (BP) neural network for classification, leveraging multi-feature data to improve accuracy. Experimental results demonstrate the effectiveness of the proposed approach, achieving a classification accuracy of 89.5%, which is significantly higher than traditional methods such as Random Forest (81.2%) and Support Vector Machine (84.7%). Furthermore, the proposed model shows consistent improvements in recall and precision across various emotion categories. These findings validate the algorithm’s robustness and practicality, offering a valuable contribution to the field of music emotion recognition.