Research On The Forecast Of Yu'ebao's Return Rate Based On The Combination Model

With the rapid development of Internet technology,Internet finance has constantly launched innovative products,which have attracted great attention from all sectors of society and exerted a great influence on traditional commercial Banks,bringing the successful experience of Internet finance to traditional finance and revolutionizing the face of traditional finance.The Internet money fund represented by Yu Ebao,relying on the alipay platform,has grown into the largest money fund and public offering fund in China in just a few years.Therefore,it is of great significance to predict the fluctuation and trend of Yu Ebao's yield data.In this paper,the seven-day annualized rate of return of Yu Ebao is taken as the research object,and a variety of combination models are established to predict and study it.Firstly,the EMD-BP combination model was established to predict and analyze Yu Ebao's annualized rate,and the prediction results were compared with the single BP neural network model.Then,based on the KNN model,the sliding window is used to extract data dynamically,and the principal component analysis is used as the method of dimension reduction and feature extraction to predict the combined model,and the PCA-KNN combined model and EMD-PCA-KNN combined model are proposed.Finally,by combining EMD-PCA-KNN with BP neural network,and taking EMD-PCA-KNN as the dimension reduction and feature extraction method,KNN as the clustering method and BP neural network as the prediction method,the BP neural network is improved and optimized,and the return rate data of Yu Ebao is predicted.Finally,through the comparison between the predicted value and the real value,the empirical analysis shows that the combined model is better than the BP neural network model.