| The popular trend of music can be predicted according to the current pop artists.So predicting the popular trend of music can be treated as predicting who the pop artist will be in the next period of time.And whether an artist is the current pop artist needs to be judged according to the auditions of artist’s music in the most recent time.Through making a statistical analysis for the records of songs including audition,download,collection,we can predict the music’s auditions of the artists in the next stage.So that we can find the artists whose music’s auditions are highest in the next period of time.It is these artists that represent the trend of music.In this thesis,the records of auditing,downloading and collecting of music are analyzed statistically and we predicted the music’s auditions of the artists using the secondary exponential smoothing method,autoregressive moving average model and BP neural network model.Futhermore,a prediction system of popular music trend based on BP neural network algorithm is presented.The main contributions of this thesis is as follows:1.By reading a large number of domestic and foreign literature,the research status of some areas are studied including prediction of music audition at home and abroad,neural network algorithm,the characteristics of the algorithm and its applications in many applications fields.With the application of BP neural network algorithm,the basic data sets to explore the factors influencing the music’s auditions of the artists are summarized and analyzed.Finally,the BP neural network algorithm is used to predict daily the music’s auditions of the artists in the next month.2.The music’s auditions of the artists is predicted based on the BP neural network algorithm.At the same time,the secondary exponential smoothing method and the autoregressive moving average model are used to predict the music ’s auditions of the artists.Finally,the prediction results of this three methods are compared.3.The prediction system of popular music trend based on BP neural network algorithm has been designed and realized.The system is developed based on J2EE platform,combined with web development technology,database technology and data mining technology.It enables those people who don’t understand the two exponential smoothing method,autoregressive moving average model,BP neural network algorithm to use the system to predict the music’s auditions of the artists in the next period of time so that the artists who will represent the popular trend of music in the next period of time could be found.Finally,this thesis summarizes the whole research content and gives some work prospects in future. |
