Error correction method based on data transformational GM(1,1) and application on tax forecasting

This paper provides an overview of the existing prediction methods, and then put forward a new approach for the improvement of prediction accuracy by introducing error correction. The characteristics of this method are to make full use of some valuable information contained in the error value. Its main idea is that the accelerated translation transforms and weighting mean value generating transforms are introduced to deal with prediction error sequence, so as to analyze the distribution of the prediction error and construct the error correction method based on data transformational GM(1,1),which is used to predict the subsequent error. And then, the predictive value of error is used to adjust the initial predicted revenue. After data transformation, the new sequences are much less volatile than before, and the smooth degree also becomes small. And to be more credible, case study verifies that this method has a better fitting effect and higher prediction precision.