A TIME SERIES FORECASTING MODEL BASED ON LINGUISTIC FORECASTING RULES

The fuzzy time series (FTS) forecasting models have been being studied intensively over the past few years. Most of researches focus on improving effectiveness FTS using time-invariant logical relationship groups proposed by Chen et al. In contrast to Chen’s model, a set can be repeated in right-hand side Yu’s model. N. C. Dieu enhanced model time-variant instead ones. mentioned above partition historical data into subintervals and assign sets them human expert’s experience. D. Hieu linguistic utilizing hedge algebras quantification converse numerical series. Similar obtained define linguistic, relationships which are used establish form this paper, we propose based rules induced from Hieu. experimental studies enrollments University Alabama observed 1971 1992 daily average temperature June 1996 September Taipei show outperformance counterpart