Perbandingan Uji Akurasi Fuzzy Time Series Model Cheng Dan Lee Dalam Memprediksi Perkembangan Harga Cabai Rawit

Abstract

Fuzzy Time Series is a method used to predict data. Fuzzy Time Series is a development of time series analysis, where Fuzzy Time Series uses the concept of fuzzy sets as the basis for its calculations. In addition, Fuzzy Time Series has various methods such as Cheng and Lee Fuzzy Time Series. In this study, Fuzzy Time Series is used to predict data on the price development of cayenne pepper in Indonesia. By using these two methods, an analysis of the level of accuracy is then carried out using several methods. So that the results obtained in this study are the MAE value of the Cheng method 669,162 and the Lee method 502,285, the MSE value of the Cheng method 1.261.393 and the Lee method 699.030.1, the MPE value of the Cheng method 0,01% and the Lee method -0,02%, and The MAPE value of the Cheng method is 1,24% and the Lee method is 0.92%. The Lee method has a smaller error value than the Cheng method, so that the Lee method is declared to be better than the Cheng method.