The Appropriate Weight Fuzzy Time Series for the Stationary Data

This book presents the appropriate weight for forecasting of AR(1) process by using fuzzy time series concept. A determination of weight approach is based on left and right (LAR) relationship using a collection of variation of chronological number in a fuzzy logical group (FLG). In the forecasting rule, the weight can be attempted into two proposed methods, namely non-reversal and reversal methods. By using data are generated from the AR(1) model and simulation technique both methods have been compared respectively. The results show that average of mean square error (MSE) from non-reversal method is smaller than reversal method on forecasting of AR(1) process. Thus, both of methods can be considered for AR(1) process. In the end of this book, the proposed method can be trained and tested by using real data