The Study of Emotional Music Classification Platform Using Artificial Neural Network Model

• Main Authors: Chung-Yi Yeh, 葉忠義
• Other Authors: Shuenn-Tsong Young
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/22524523339090752221

碩士 === 國立陽明大學 === 醫學工程研究所 === 98 === Music material is the crucial factor for the application of emotion guidance or the induction of positive mood in the music therapy. Understanding the music structure or composition is helpful to classify different emotional music that can be applied in music therapy. Traditionally, people usually classified the emotion of music according to the experience of music specialists, but this classification method consumes time and the manpower heavily. The present study aimed to establish an automatic classifier with artificial neural networks for the classification of emotional music. The results of this study can be applied to simplify the creation and update of music database for musical therapy. Three methods were adopted to improve the performance of neural network with back propagation algorithm. Summary, the classifier can classify four types of emotional music (Tension, Gaiety, Gloom and Relaxation) with accuracy rate of 81.5%. The classified music had the musical characteristics agreeing with the music theory.