Text
Peramalan Menggunakan Metode Weighted Fuzzy Integrated Time Series (Studi Kasus: Harga Beras di Indonesia Bulan Januari 2011 s/d Desember 2017)
ABSTRAK
Fuzzy Time Series (FTS) merupakan teknik peramalan data runtun waktu yang menggunakan konsep teori fuzzy. Sistem peramalan menggunakan FTS berguna untuk menangkap pola data masa lalu kemudian digunakan untuk menghasilkan informasi di waktu yang akan datang. Awalnya dalam FTS setiap pola relasi yang terbentuk dianggap memiliki bobot yang sama selain itu hanya menggunakan orde pertama, pada perkembangannya muncul Weighted Fuzzy Integrated Time Series (WFITS) yang memberikan pembedaan bobot setiap relasi dan penggunaan orde tinggi. Pengukuran ketepatan hasil peramalan digunakan nilai Root Mean Square Error (RMSE) dan Mean Absolute Percentage Error (MAPE). Pada penelitian ini metode WFITS model orde pertama maupun orde tinggi diterapkan untuk meramalkan harga beras di Indonesia berdasarkan data bulan Januari 2011–Desember 2017. Dari hasil analisis diperoleh peramalan data menggunakan WFITS algoritma Lee model orde tinggi (1,2,3) memberikan nilai RMSE dan MAPE pada data testing secara berturut-turut sebesar 69,898 dan 0,47%, untuk nilai RMSE dan MAPE pada data training secara berturut-turut sebesar 70,4039 dan 0,54%.
Kata Kunci : Fuzzy Time Series, Weighted Fuzzy Integrated Time Series, RMSE,
MAPE, Orde Tinggi, Harga Beras
ABSTRACT
Fuzzy Time Series (FTS) is a forecasting technique that uses fuzzy theory concepts. Prediction systems using FTS are useful for capturing past data patterns and then used to produce information in the future. Initially in the FTS each pattern of relationships formed was considered to have the same weight. Besides using only a single order, there are now many developments from FTS, one of which is the Weighted Fuzzy Integrated Time Series (WFITS) which gives a weight differentiation for each relation and high order usage. To see the accuracy of the forecasting results of the data seen by calculating the level of accuracy or measure of forecasting accuracy by using the Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). In this study the WFITS method of the first-order and first-order models was applied to forecast rice prices in Indonesia from January 2011 - December 2017 with a measure of accuracy using RMSE and MAPE. From the analysis results obtained that data prediction using Lee's high order model WFITS algorithm (1,2,3) provides the best accuracy with RMSE value of 69,898 and 0.47% MAPE value for data testing and RMSE value of 70,4039 and MAPE value 0.54% for training data.
Keywords: Fuzzy Time Series, Weighted Fuzzy Integrated Time Series, RMSE,
MAPE, High Order, Rice Prices
703E19II | 703 E 19 | Perpustakaan FSM Undip (Referensi) | Tersedia |
Tidak tersedia versi lain