Perbandingan Metode Fuzzy Time Series dan Holt Double Exponential Smoothing Pada Peramalan Jumlah Mahasiswa Baru Institut Pertanian Bogor

Abstract

Forecasting is an activity to predict values of a variable in the future. The purposes of this research is predicting the number of new students at Bogor Agriculture University using fuzzy time series method and Holt double exponential smoothing method. This research is also comparing these two methods to find the more appropriate method based on the accuracy of forecasting showed by Mean Absolute Percentage Error (MAPE). Fuzzy time series methods used fuzzy set in the process of forecasting while Holt double exponential smoothing method smooths the value of serial data by reducing them exponentially. In forecasting the number of new students at Bogor Agricultural University, fuzzy time series method, with MAPE as big as 6.412%, shows better forecasting accuracy than Holt double exponential smoothing method which has MAPE as big as 7.75%. After conducting a case study, we conclude that Holt double exponential smoothing method will be more accurate if there are many data available.