A Research On The Model Of Music Emotion Based On The Neural Network

The analysis on musical emotion is one research direction of artificial intelligence. The research aims to make the computer identify the emotion of music. Currently, the research on the analysis of music emotion is mainly utilized in the automatioc composition by computers and the search for music which bases on emotion.Based on the need of psychology and neural network modelling, the thesis establishes a new emotion classification model on a basis of an old emotion classification; it puts forward a new algorithm based on pitch continuity and note intermittence, and realizes the location of the main music track; it designs the algorithm in order to select information from the main music track; it makes a reference to the representation of music characteristics in music theory, and makes a mathematical description of music emotion by analyzing music information; finally it designs and testifies the model of neural network based on the BP algorithm in practice.