Song and Chissom first proposed the fuzzy time series forecasting model in 1993. In this paper, we improved the forecasting model proposed by Stevenson and Porter, and dug out the SIFBODR (The Set of Improved Fuzzy Time Series Forecasting Models Based on the Ordered Difference Rate). In the research on the forecasting problem of enrollments of the University of Alabama 1971-1992, the forecasting model SIFBODR(0.00002, 0.00004) of SIFBODR can obtain AFER (Average Forecasting Error Rate) = 0% and

more »

... MSE(Mean Square Error) = 0. The problem that the prediction accuracy of fuzzy time series forecasting models is not high for many years is basically solved. Keywords-fuzzy time series forecasting method; fuzzy number function of SIFBODR; inverse fuzzy number function of SIFBODR; forecasting function of SIFBODR I. R