Ali, Noor Rasidah and Ku-Mahamud, Ku Ruhana (2018) High-order RTV-FUZZY time series forecasting model based on trend variation. Journal of Theoretical and Applied Information Technology, 96 (21). pp. 7151-7163. ISSN 1992-8645

PDF Restricted to Registered users only Download (279kB) | Request a copy

Abstract

Time series data principally involves four major components which are trend, cyclical, seasonal and irregular, that reflects the characteristics of the data. Ignoring the systematic analysis of patterns from time series components will affect forecasting accuracy. Thus, this paper proposes a high-order ratio trend variation (RTV) fuzzy time series model based on the trend pattern and variations in time series to deal with patterns within the time series data. RTV is used in the fuzzification process to deal with data that contains vagueness, uncertainty and impreciseness. Proper adjustment was also applied to handle the common issues in fuzzy time series model includes determination of length of interval, fuzzy logic relations (FLRs), assigning weight to each FLR and the defuzzification process. Empirical analysis was performed on enrollments data of Alabama University to assess the efficiency of the model. The performance of the proposed model was evaluated by comparing the average forecasting error rate and mean square error values with several fuzzy time series models in the literatures. Experimental results revealed that the proposed model was better than other fuzzy time series models. The use of RTV was able to grip the fuzziness in time series data and reduce the estimation of forecasting errors. In addition, this technique is capable to identify and describe the underlying structure that influence the occurrence of the uncertainty and high fluctuation of the phenomena under investigation.

Item Type:	Article
Uncontrolled Keywords:	High-Order Fuzzy Time Series, Ratio Trend Variation, Enrolment, Fuzzy Logic Relation
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:	School of Computing
Depositing User:	Mrs. Norazmilah Yaakub
Date Deposited:	11 Nov 2020 05:56
Last Modified:	11 Nov 2020 05:56
URI:	https://repo.uum.edu.my/id/eprint/27871

Actions (login required)

View Item