基于逆模糊数的模糊时间序列预测的新方法  

New Method of Fuzzy Time Series Forecasting Based on Inverse Fuzzy Number

作　　者：王鸿绪[1] 陈德广[1] 过建春[1] 金海龙[1] 

机构地区：[1]琼州学院旅游管理学院,海南三亚572022

出　　处：《数学的实践与认识》2015年第7期240-248,共9页Mathematics in Practice and Theory

基　　金：琼州学院博士启动基金(QYXB201402);琼州学院旅游管理学院省级特色专业海南省第三批重点(扶持)学科建设项目

摘　　要：Song和Chisson于1993年提出模糊时间序列预测理论．虽然至今已经提出许多模糊时间序列预测模型，但是迄今为止尚未给出预测未知年数据的模糊时间序列预测模型．提出基于逆模糊数的模糊时间序列预测的新方法该模型对广西大学已知的2001-2012年的注册数进行预测分析，平均预报误差率比相关文献的方法有所改善．还对广西大学未知的2013年和2014年注册数进行预测研究．方法是短期预测的一种模糊时间序列预测方法．Song and Chisson proposed the fuzzy time series forecasting theory in 1993. Though to this day many fuzzy time series forecasting model have been proposed, but so far the fuzzy time series prediction model for predicting the data in unknown years is not provided. This paper presents a new method of fuzzy time series forecasting based on inverse fuzzy numbers. The model about the enrollments of the known 2001-2012 years of Guangxi University to be in progress forecasting research, the average forecasting error rate than the method of reference [9] has improved About the enrollments of the unknown 2013 year and 2014 year of Guang-xi University to be also in progress forecasting research. This method is a kind of fuzzy time series forecasting method for short-term prediction.

关 键 词：模糊集 时间序列 预测方法 注册数的差 逆模糊数 

分 类 号：O159[理学—数学]