| With the continuous development and progress of science and technology,artificial intelligence has penetrated into all walks of life.In the current era of data and information,AI technology and artificial intelligence are unavoidable topics in every industry.As contemporary music artists,how to make good use of big data to work more efficiently and quickly on the basis of traditional music research is a problem worth thinking about.Artificial intelligence and music have many consistent research directions.This paper chooses AI composition as the main object of discussion and research,and intends to discuss the feasibility of artificial intelligence in the field of AI composition by illustrating and exploring practical methods,so as to look into the development prospect of artificial intelligence in music technology in the future.In this paper,I use Keras deep learning framework for AI composition practice as the fundamental analysis of the thesis,from the perspective of theory and practice to demonstrate the possibility of AI composition.In the process of the experiment,I chose Python as my basic programming language and designed and experimentally with Keras framework.During the experiment,I gave the artificial intelligence network a certain number of experimental materials of music MIDI files,and made the artificial intelligence train and generate the neural network model according to the music materials I provided through the neural network,and finally generated and created the music according to the generated model.At the end of the experiment,I successfully made use of neural network to generate original music works by artificial intelligence.I take three music materials trained by artificial intelligence as the analysis object,longitudinal comparison between them and the original music materials,starting from the aspects of artistry and logic,analysis of artistry and feasibility of artificial intelligence music creation. |
