Title: Applications of axiomatic fuzzy sets theory on fuzzy time series forecasting

Authors: Angelo Dan Menga Ebonzo; Xiaodong Liu

Addresses: Transportation Management College, Dalian Maritime University, No. 1 Lingnanzhong Road, High-tech Zone, Dalian, 116026, China. ' Department of Mathematics, Dalian Maritime University, Dalian 116026, China

Abstract: A number of methods have been proposed for forecasting based on fuzzy time series. Most of fuzzy time series are presented for forecasting enrolments. In this paper, we propose an innovate fuzzy time series forecasting model using axiomatic fuzzy set (AFS) theory. The advantages of using AFS theory in this approach are multiple: fuzzy sets with more than one maximum value are obtained, this affects considerably the forecasting accuracy; values of membership degrees are directly obtained from the data. Compared with existing methods, the experimental study shows that the proposed method can get best forecasting accuracy rate over methods described in the literature.

Keywords: fuzzy time series; enrolments; time series forecasting; fuzzy set theory; linguistic variable; fuzzy logic relationships; FLRs; axiomatic fuzzy sets; fuzzy forecasting; forecasting accuracy.

DOI: 10.1504/IJSCC.2012.050824

International Journal of Systems, Control and Communications, 2012 Vol.4 No.4, pp.280 - 295

Received: 09 Feb 2012
Accepted: 14 Jun 2012

Published online: 04 Dec 2012 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article