Key points are not available for this paper at this time.
With the rapid developemnt of machine learning techiniques, it has been widely adopyed in various fields. This paper explores the models and applications of machine learning in music composition. It traces the evolution of computer music from its early days with electronic music pioneers to its contemporary implementations, showcasing the integration of cutting-edge technologies. Various models and techniques employed in music composition, including rule-based systems, deep learning, and stochastic composition methods, are thoroughly examined. Furthermore, significant challenges posed by copyright issues are addressed, along with potential solutions to overcome them. Looking ahead, one envisions a future where computer music achieves greater emotional depth, accessibility, and customization despite inherent complexities. Advances in copyright management and licensing systems are expected to diversify the pool of musical compositions available for training machine learning models, resulting in more emotionally resonant and diverse computer-generated music. This study sheds light on the potential of combining computer music with machine learning, providing valuable insights into the intricate world of music composition and technological innovation for future music creators and enthusiasts.
Xiaohan Yang (Wed,) studied this question.