Open AccessPengujian Metode Fuzzy Time Series Chen dan Hsu Untuk Meramalkan Nilai Indeks Bursa Saham Syariah Di Jakarta Islamic Index (JII)
Author(s) -
Rizka Zulfikar,
Prihatini Ade Mayvita
Publication year - 2018
Publication title -
wiga
Language(s) - English
Resource type - Journals
eISSN - 2549-5992
pISSN - 2088-0944
DOI - 10.30741/wiga.v7i2.340
Subject(s) - mathematics , stock market index , chen , statistics , series (stratigraphy) , stock market , index (typography) , econometrics , computer science , geography , paleontology , context (archaeology) , archaeology , world wide web , biology
This research is an empirical study to tested the accuracy of Chen and Hsu’s Fuzzy Time Series Method used to forecast sharia market stock index in Jakarta Islamic Index. The data used in this research are secondary data consists of daily stock market indexes during 23 November 2016 to 14
July 2017. Chen dan Hsu’s Fuzzied Series Method used in this research has the smallest MSE (Mean Square Error) and AFER (Average Forecasting Error Rate) value rather than others method such as Song and Chrissom (1993). Song and Chrissom (1994), Chen (1996), Hwang, Chen and Lee (1998),
Huarng (2001) and Chen (2002).
To tested the accuracy of the Chen’s dan Hsu’s Fuzzied Series
Method researcher has to do 5 (five) steps such as (1) Determine lag between historical data, interval and The Universe Data (U), (2) Distributing Data into The Unniverse, (3) Define The Fuzzy Set, (4) Determine The Fuzzy Logical Relationship (FLR), and (5) Analyse the Difference between data. There are 3 (three) rules in Chen’s dan Hsu’s Fuzzied Series Method based on the Difference and FLR. The result of this research is Chen dan Hsu’s Fuzzied Series Method has MSE = 1.88 and AFER =0.006% and it can be used to make forecasting on value and trend sharia stock market in Jakarta Islamic index.