Research on the efficacy of music aesthetic education on college students’ self-identity enhancement based on neural network modeling

In this paper, the entire chord progression is added to the generation process through a bidirectional LSTM model, and the Skip-connection method is used to accelerate the convergence speed in all recursive layers except the first one. Different musical emotions are classified based on the Hevner emotion model, and features such as pitch, duration, and tempo of musical emotions are parameterized. The forward neural network is used to construct the music emotion classification model, and the gradient descent learning algorithm is used to algorithmically control the forward neural network model. At the same time, this paper explains the significant enhancement of college students’ self-identity by music aesthetic education based on neural network model from two perspectives: theoretical research and empirical analysis. The results show that the music generation and music emotion classification models constructed based on the neural network algorithm in this paper show good performance in the experiments. After applying the neural network model containing music generation and music emotion classification to music aesthetic education and counseling college students on self-identity, the mean score of self-identity scale of students in the experimental group increased from 50.83 to 88.56, with an improvement of 75.78%, and the results were significant at the 1% level. The effectiveness of this paper’s method in enhancing college students’ self-identity is fully demonstrated.