Design of an Improvisational Singing Training System Based on Machine Learning Algorithms
摘要
There is a fundamental difference between improvisation and traditional performance at the source of the music played. The traditional method is to start from the outside, which is to play based on the notes recorded in the music. “Improvisation” is a form of performance that comes from the depths of people’s hearts and does not strive to achieve perfection. It focuses more on the performer’s inner world. This article first provided an overview of the definition, classification, development history, principles, and evaluation criteria for improvisation skills of machine learning algorithms. Finally, through experimental analysis, it can be concluded that the student scores based on this algorithm have improved in the evaluation of improvisation skills, fully verifying the practical significance of the algorithm.