PERBANDINGAN METODE PEMULUSAN EKSPONENSIAL TUNGGAL DAN FUZZY TIME SERIES UNTUK MEMPREDIKSI INDEKS HARGA SAHAM GABUNGAN

Abstract

The development of methods of forecasting with time series data quite rapidly result there are many options that the method can be used to predict the data according to the needs and the need to compare one method to the other methods that get results of prediction with high accuracy. In this thesis, comparison of forecasting will be done using measure forecasting accuracy in the form of MAPE, MAE, and MSE of a forecast in calculating the value of The composite stock price index (CSPI) using Single Exponential Smoothing method that will be compared to modern forecasting methods, namely Fuzzy Time Series . Fuzzy Time Series methods used in this study is the method of Fuzzy Time Series proposed by Chen and Cheng. Between the three forecasting methods obtained the best method is of Cheng’s Fuzzy Time Series. Keywords: Method of Forecasting, Single Exponential Smoothing, Fuzzy Time Series, Chen’s and Cheng’s Method, CSPI