Interval Based in Fuzzy Sliding Window for Forecasting Crude Palm Oil

Nur Fazliana, Rahim and Mahmod, Othman and Rajalingam, Sokkalingam (2020) Interval Based in Fuzzy Sliding Window for Forecasting Crude Palm Oil. Menemui Matematik (Discovering Mathematics), 42 (2). pp. 96-104. ISSN 2231-7023

[img] PDF
Interval Based in Fuzzy Sliding Window for Forecasting Crude Palm Oil.pdf

Download (506kB)

Abstract

Interval is the main component in time series forecasting, hence a Fuzzy Sliding Window Forecasting Method (SWM) suggested in obtaining intervals of forecasting in the Fuzzy Time Series (FTS). Formerly, almost all the intervals were calculated using class frequency. The intervals are then regrouping into the sub-intervals using the provided category. Whereas in this study, the prediction of interval obtained by embedding the idea of SWM into FTS forecasting. The intention of this suggested method is to further improve the success of a time series forecast and indirectly increase forecasting precision. The daily prices of Crude Palm Oil (CPO) data are taken for verification purposes. Hence, the precision of the suggested method is differentiating the existing forecasting method. The outcome of this method is compared to the other methods and it reveals that the suggested method produces precise intervals determination. The discovery of this study can be used as a replacement of existing forecasting method to get an improved prediction interval.

Item Type: Article
Uncontrolled Keywords: Forecasting, Fuzzy Time Series, Intervals, Sliding Window
Subjects: Q Science > QA Mathematics
Divisions: Academic Faculties, Institutes and Centres > Centre for Pre-University Studies
Faculties, Institutes, Centres > Centre for Pre-University Studies
Academic Faculties, Institutes and Centres > Centre for Pre-University Studies
Depositing User: Tuah
Date Deposited: 29 Jan 2021 08:17
Last Modified: 29 Jan 2021 08:17
URI: http://ir.unimas.my/id/eprint/34154

Actions (For repository members only: login required)

View Item View Item