Home > Research > Publications & Outputs > Fuzzy time series forecasting with a novel hybr...


Text available via DOI:

View graph of relations

Fuzzy time series forecasting with a novel hybrid approach combining fuzzy c-means and neural networks

Research output: Contribution to Journal/MagazineJournal articlepeer-review

<mark>Journal publication date</mark>15/02/2013
<mark>Journal</mark>Expert Systems with Applications
Issue number3
Number of pages4
Pages (from-to)854-857
Publication StatusPublished
<mark>Original language</mark>English


In recent years, time series forecasting studies in which fuzzy time series approach is utilized have got more attentions. Various soft computing techniques such as fuzzy clustering, artificial neural networks and genetic algorithms have been used in fuzzy time series method to improve the method. While fuzzy clustering and genetic algorithms are being used for fuzzification, artificial neural networks method is being preferred for using in defining fuzzy relationships. In this study, a hybrid fuzzy time series approach is proposed to reach more accurate forecasts. In the proposed hybrid approach, fuzzy c-means clustering method and artificial neural networks are employed for fuzzification and defining fuzzy relationships, respectively. The enrollment data of University of Alabama is forecasted by using both the proposed method and the other fuzzy time series approaches. As a result of comparison, it is seen that the most accurate forecasts are obtained when the proposed hybrid fuzzy time series approach is used.