سال انتشار: ۱۳۹۱

محل انتشار: چهارمین کنفرانس مهندسی برق و الکترونیک ایران

تعداد صفحات: ۵

نویسنده(ها):

Narges Shafaei – Islamic Azad University, Gonabad Branch Gonabad, Iran
Asef Zare – Islamic Azad University, Gonabad BranchGonabad, Iran

چکیده:

In this paper, Taiwan stock index is forecasted byuse of adaptive type 2 fuzzy time series concept. In Type 2fuzzy time series, more observation can be used for forecasting.Therefore, comparing with type 1 model not only this model iscloser to reality but also the results of forecasting with thismodel are more than effective. In addition, using adaptivewindow size makes higher average forecasting accuracy ratesthan the previous methods.