Prediksi Populasi Penduduk di Indonesia Menggunakan Prophet dan SVR

Libriawan, Raditya Dimas (2025) Prediksi Populasi Penduduk di Indonesia Menggunakan Prophet dan SVR. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) Cover.pdf Download (2MB)
	Text (Bab 1) Bab 1.pdf Download (225kB)
	Text (Bab 2) Bab 2.pdf Restricted to Repository staff only until 3 December 2027. Download (538kB)
	Text (Bab 3) Bab 3.pdf Restricted to Repository staff only until 3 December 2027. Download (582kB)
	Text (Bab 4) Bab 4.pdf Restricted to Repository staff only until 3 December 2027. Download (1MB)
	Text (Bab 5) Bab 5.pdf Download (287kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf Download (183kB)
	Text (Lampiran) Lampiran.pdf Restricted to Repository staff only until 3 December 2027. Download (4MB)

Abstract

The population size of a country is a very important aspect because it directly affects various aspects of life. Indonesia ranks fourth as the country with the largest population in the world. Indonesia's high population brings serious problems to life. Therefore, monitoring and controlling population growth is a crucial and inseparable step, one of which is by utilizing machine learning to perform time series data forecasting. This study was conducted to determine the performance of the Facebook Prophet (Prophet) and Support Vector regression (SVR) algorithms in predicting the population of Indonesia. The test results show that the combination of the Prophet SVR model is able to provide the best performance compared to tests on other comparative models. The Prophet-SVR model testing results obtained accurate evaluation values with MAE of 0,27×107, RMSE of 0,29×107, and MAPE of 1.04%. These values tend to be lower when compared to the results of testing the Prophet model alone, proving that SVR is able to correct the Prophet model results by capturing non-linear patterns in the residual data. The best test results were obtained with a data split ratio of 85% training data and 15% test data, with hyperparameter configurations of 0.5 for CPS; 15 for n-changepoints; 0.1 for SPS; ‘multiplicative’ seasonal mode; SVR model parameter C value with Random Search optimization of 90; epsilon of 0.5; ‘rbf’ kernel; and ‘scale’ mode on gamma. This study confirms that for complex demographic data, the hybrid model approach has proven to be significantly superior to single models, resulting in more accurate forecasting solutions for long-term planning.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Sari, Anggraini Puspita	NIDN0716088605	anggraini.puspita.if@upnjatim.ac.id
Thesis advisor	Wahanani, Henni Endah	NIDN0022097811	henniendah.if@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Raditya Dimas Libriawan

Date Deposited:

04 Dec 2025 04:11

Last Modified:

04 Dec 2025 04:11

URI:

https://repository.upnjatim.ac.id/id/eprint/47677