Efendi, Riswan and Mat Deris, Mustafa (2018) Non-Probabilistic Inverse Fuzzy Model in Time Series Forecasting. International Journal of Uncertainty, Fuzziness and Knowledge Based Systems, 26 (05). pp. 855-873. ISSN ISSN (print): 0218-4885 | ISSN (online): 1793-6411

Preview

Text Non-Probabilistic Inverse Fuzzy Model in Time Series Forecasting.pdf - Published Version Download (1MB) | Preview

Abstract

Many models and techniques have been proposed by researchers to improve forecasting accuracy using fuzzy time series. However, very few studies have tackled problems that involve inverse fuzzy function into fuzzy time series forecasting. In this paper, we modify inverse fuzzy function by considering new factor value in establishing the forecasting model without any probabilistic approaches. The proposed model was evaluated by comparing its performance with inverse and non�inverse fuzzy time series models in forecasting the yearly enrollment data of several universities, such as Alabama University, Universiti Teknologi Malaysia (UTM), and QiongZhou University; the yearly car accidents in Belgium; and the monthly Turkish spot gold price. The results suggest that the proposed model has potential to improve the forecasting accuracy compared to the existing inverse and non-inverse fuzzy time series models. This paper contributes to providing the better future forecast values using the systematic rules. Keywords: Fuzzy time series, inverse fuzzy function, non-probabilistic model, non-inverse fuzzy model, future forecast

Actions (login required)

View Item