Andalita, Ilafi (2015) Peramalan jumlah penumpang kereta api kelas ekonomi Kertajaya menggunakan ARIMA dan ANFIS. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
1311100038-Undergraduate_Thesis.pdf Download (2MB) | Preview |
Preview |
Text
1311100038-Presentation-presentationpdf.pdf Download (2MB) | Preview |
Preview |
Text
1311100038-Paper-1311100038-paperpdf.pdf Download (735kB) | Preview |
Abstract
Peramalan jumlah penumpang kereta api kelas ekonomi Kertajaya merupakan salah satu upaya penting untuk mengetahui kebutuhan pengguna transportasi tersebut. Namun, time series jumlah penumpang kereta api Kertajaya yang mempunyai fluk-tuasi tinggi terbukti bersifat nonlinear berdasarkan uji linieritas. Pemodelan menggunakan pendekatan linier seperti ARIMA Box-Jenkins tidak selalu memberikan peramalan yang baik karena ter-ikat beberapa asumsi dalam membangun model. Oleh karena itu penelitian ini mengusulkan penggunaan metode ANFIS yang diharapkan memberikan kinerja lebih baik dalam pemodelan nonlinier dan dibandingkan dengan hasil dari ARIMA. Penggu-naan ANFIS untuk peramalan jumlah penumpang selama 14 peri-ode ke depan memberikan akurasi ramalan yang lebih tinggi da-ripada ARIMA karena MAPE dan RMSE yang dihasilkan lebih kecil. Model ANFIS terbaik dihasilkan dari input jumlah penum-pang pada satu, tujuh, dan delapan hari sebelumnya dan fungsi keanggotaan pi.
==============================================================================================================
Forecasting the passenger volume of economy class train Kertajaya is essential to know the needs of transport users. However, time series of passenger volume from Kertajaya train which has a high fluctuation is proved to be nonlinear based on linearity test. Modeling using linear approach such as Box-Jen-kins ARIMA model not always performs better because it’s bound some assumptions to build the model. Therefore, this study proposes ANFIS model that is expected can yields better per-formance in modeling nonlinear time series and compared with the results of ARIMA. Using ANFIS for forecasting the passenger volume during the 14 periods aheads yields a higher accuracy of forecasts than ARIMA because it gives smaller MAPE and RMSE. The best ANFIS model is generated from the input of passenger volume on one day, seven and eight days before and the mem-bership function of pi.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSt 519.535 And p |
| Uncontrolled Keywords: | ARIMA, ANFIS, jumlah penumpang, nonlinier |
| Subjects: | H Social Sciences > HB Economic Theory > Economic forecasting--Mathematical models. |
| Divisions: | Faculty of Mathematics, Computation, and Data Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | - Taufiq Rahmanu |
| Date Deposited: | 04 Dec 2019 06:51 |
| Last Modified: | 04 Dec 2019 06:51 |
| URI: | http://repository.its.ac.id/id/eprint/72194 |
Actions (login required)
 |
View Item |