Penerapan Metode Maximal Overlap Dicrete Wavalet Transform Sarima Untuk Peramalan Banyaknya Penumpang Kereta Api Jabodetabek (Berdasarkan Data Periode Januari 2015 – April 2021)

Amalya Fatonah, - (2021) Penerapan Metode Maximal Overlap Dicrete Wavalet Transform Sarima Untuk Peramalan Banyaknya Penumpang Kereta Api Jabodetabek (Berdasarkan Data Periode Januari 2015 – April 2021). S1 thesis, Universitas Pendidikan Indonesia.

Baca Full Text klik disini
[img] Text
S_MAT_1701069_Title.pdf
Download (298kB)
[img] Text
S_MAT_1701069_Chapter1.pdf
Download (152kB)
[img] Text
S_MAT_1701069_Chapter2.pdf
Restricted to Staf Perpustakaan
Download (395kB)
[img] Text
S_MAT_1701069_Chapter3.pdf
Download (193kB)
[img] Text
S_MAT_1701069_Chapter4.pdf
Restricted to Staf Perpustakaan
Download (1MB)
[img] Text
S_MAT_1701069_Chapter5.pdf
Download (86kB)
[img] Text
S_MAT_1701069_Appendix.pdf
Restricted to Staf Perpustakaan
Download (641kB)
Official URL: http://repository.upi.edu

Abstract

The use of wavelet decomposition for statistical modeling, especially in time series data, has experienced rapid development. MODWT-SARIMA is considered more suitable for time series data because in each level of decomposition there is a wavelet coefficient and a scale along the length of the data. This study aims to predict the number of Jabotabek train passengers based on time series data and prove that non-stationary Jabodetabek train passenger time series modeling is more effective than SARIMA alone.MODWT-SARIMA which is a time series modeling method that combines the MODWT process and the SARIMA process. The MODWT process is used as the pre-processing of the data, while the SARIMA process is used as the model formation. The results of the SARIMA model diagnostic check for MODWT decomposition data, namely D1,D2¬, D3 and S3 show that the residual model is not white noise while SARIMA is already white noise. Theoretically, a model that is not white noise is less able to describe the properties of the observed data because it contains information that needs to be considered. However, this study of the MODWT-SARIMA model has shown that the MODWT-SARIMA model is more effective for modeling time series which is not stationary compared to the SARIMA model. The accuracy of the method is seen from the forecast results which are based on the RMSE value. The results of the MODWT-SARIMA model have a value of RMSE = 10286.71 which is smaller than the SARIMA model with RMSE = 11334.02. This shows that the MODWT-SARIMA method is more effective for non-stationary data compared to SARIMA. Keywords : MODWT-SARIMA, time series, train passengers.

Item Type: Thesis (S1)
Uncontrolled Keywords: MODWT-SARIMA, runtun waktu, penumpang kereta api
Subjects: L Education > L Education (General)
Q Science > QA Mathematics
Divisions: Fakultas Pendidikan Matematika dan Ilmu Pengetahuan Alam > Jurusan Pendidikan Matematika > Program Studi Matematika (non kependidikan)
Depositing User: Amalya Fatonah
Date Deposited: 30 Aug 2021 07:15
Last Modified: 30 Aug 2021 07:15
URI: http://repository.upi.edu/id/eprint/64614

Actions (login required)

View Item View Item