Performance prediction of Chinese family business based on improved wavelet relevance vector regression
This article studies the performance of family business based on the data of Chinese listed family firms. The scientific and accurate prediction of the performance of family business has a great significance to make the strategies of company development. In the study, the improved wavelet relevance vector regression algorithm is first presented and proposed to predict the performance of Chinese family business. We collect some Chinese family firms including Baoxin energy, Etern group, Hejia group, Four dimensions controlling, Fangda group, Ingenious development as our data base. In each family business, the quarterly data from 2001-12-31 to 2011-03-31 are employed to testify the superiority of improved wavelet relevance vector regression algorithm compared with the traditional relevance vector regression algorithm. The testing results by the comparison of performance prediction of family business between improved wavelet relevance vector regression model and traditional relevance vector regression model show that the prediction ability of improved wavelet relevance vector regression model is better than that of traditional relevance vector regression model.
Author's Name: Xu, P., Cui, W.
Volume: Volume 7
Issues: Issue 15
Keywords: Business performance, Family business, Prediction algorithm, Relevance vector regression, Wavelet