PREDIKSI JUMLAH PENUMPANG KERETA API MENGGUNAKAN MODEL VARIASI KALENDER DENGAN DETEKSI OUTLIER (Studi Kasus PT. Kereta Api Indonesia DAOP IV Semarang)
| Main Authors: | Saputri, Ani Funtika, Hoyyi, Abdul, Sugito, Sugito |
|---|---|
| Format: | Article info application/pdf Journal |
| Bahasa: | eng |
| Terbitan: |
Departemen Statistika FSM Undip
, 2018
|
| Subjects: | |
| Online Access: |
https://ejournal3.undip.ac.id/index.php/gaussian/article/view/19301 https://ejournal3.undip.ac.id/index.php/gaussian/article/view/19301/18315 |
Daftar Isi:
- Transportation is an inseparable and indispensable part of society in everyday life. Trains became one of the most popular public transportation, especially during the Eid. The shifting of the lunar month of Eid forms a pattern called calendar variation. The calendar variation model is a model that combines the dummy regression model with the ARIMA model. In time series models sometimes there are outliers that can affect the suitability of the model. So that modeling and forecasting method is done using model of calendar variation with outlier detection. Based on the analysis that has been done on the data of the number of passengers of Argo Bromo Anggrek railway, we get the ARIMA model ([11], 0, 1), Dt, Dt-2,t with the addition of 4 outliers as the best model and the resulted forecasting shows increase Railway passengers increase in the months leading up to Eid. Keywords: Train, Calendar Variations, Outlier Detection