Time Series Decomposition Algorithm For Solving Music Flow Forecast

In recent years,with the improvement of traffic prediction accuracy,traffic prediction in the music song flow forecast also become more and more important.Music songs as the core of music song flow prediction is the attention and exploration of many scholars,mainly for music genre and listeners to listen to the song to identify two issues.At present,many scholars and related experts have put forward some corresponding solutions,but so far there are still many problems to be solved.In this paper,we use the operability and convergence of time series and the method of joint weighted fuzzy kernel clustering to form coding in different time periods.Based on the feasible factors Space music flow prediction algorithm,including user classification strategy,song traffic forecasting strategy.The algorithm is designed to solve the problem of song traffic prediction,and the model is constructed for the music song flow.Firstly,the music song user is modeled by using the fuzzy kernel clustering model and the convolution neural network classification model based on AdaBoost,which is used to solve the problem of pre-data classification caused by different users’preferences.Secondly,through the construction of the music song prediction model based on the user classification based on the user classification,the algorithm based on the feasible coefficient space segmentation is used to predict the use of the feasible coefficient space segmentation algorithm to solve the problem because the data fluctuation range is too large Lead to the problem of prediction bias.Finally,through the experimental analysis,the performance and efficiency of the algorithm are validated.The main innovation of this thesis includes the following:1,Effective combination of weighted fuzzy kernel algorithm and feasible coefficient space algorithm;2,Which combines the AdaBoost’s convolution neural network algorithm and the feasible coefficient space algorithm.3,the coding based feasible coefficient space algorithm applied to the music song broadcast traffic prediction solution;4,to improve the music player traffic prediction model in the artist song flow prediction efficiency and accuracy.