Prediksi Nilai Tukar Petani Di Jawa Timur Menggunakan Metode Holt’s Double Exponential Smoothing Dan Holt-Winters Dengan Optimasi Levenberg-Marquardt

Yanti, Lisa Dama (2025) Prediksi Nilai Tukar Petani Di Jawa Timur Menggunakan Metode Holt’s Double Exponential Smoothing Dan Holt-Winters Dengan Optimasi Levenberg-Marquardt. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21083010095_Cover.pdf Download (940kB)
	Text (Bab 1) 21083010095_BAB I.pdf Download (358kB)
	Text (Bab 2) 21083010095_BAB II.pdf Restricted to Registered users only until 19 June 2028. Download (627kB)
	Text (Bab 3) 21083010095_BAB III.pdf Restricted to Registered users only until 19 June 2028. Download (467kB)
	Text (Bab 4) 21083010095_BAB IV.pdf Restricted to Registered users only until 19 June 2028. Download (1MB)
	Text (Bab 5) 21083010095_BAB V.pdf Download (216kB)
	Text (Daftar Pustaka) 21083010095_DAFTAR PUSTAKA.pdf Download (212kB)
	Text (Lampiran) 21083010095_LAMPIRAN.pdf Restricted to Registered users only until 19 June 2028. Download (450kB)

Abstract

The Farmer's Exchange Rate (NTP) is an indicator of farmer welfare calculated from the ratio between the price received and production costs. In East Java Province, the NTP shows fluctuations with certain trends that reflect instability in the agricultural sector. This instability risks reducing farmers' purchasing power and disrupting production sustainability. Therefore, an accurate forecasting model is needed as a basis for policy making. This study compares two forecasting methods, namely Holt's Double Exponential Smoothing (DES) and Holt-Winters, which are both effective for trend-patterned data but differ in their handling of seasonal components. The accuracy of both methods is highly dependent on the selection of smoothing parameters, which are typically determined through trial and error. Therefore, the Levenberg-Marquardt algorithm is used to optimize parameters, as it minimizes error through a combination of the Gauss-Newton and Gradient Descent methods. The data used are monthly NTP data for East Java from 2014 to 2024 from the Central Statistics Agency (BPS). The results show that Holt DES with optimization provides the best performance with a MAPE value of 0.86%, lower than the optimized Holt-Winters (MAPE 0.94%). This indicates that Holt DES is more suitable for modeling NTP that tends to have a trend pattern without strong seasonal components.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Saputra, Wahyu Syaifullah Jauharis	NIDN0725088601	wahyu.s.j.saputra.if@upnjatim.ac.id
Thesis advisor	Damaliana, Aviolla Terza	NIDN0002089402	aviolla.terza.sada@upnjatim.ac.id

Subjects:

H Social Sciences > HA Statistics
Q Science > QA Mathematics
Q Science > QA Mathematics > QA76.6 Computer Programming

Divisions:

Faculty of Computer Science > Departemen of Data Science

Depositing User:

Lisa Dama Yanti

Date Deposited:

20 Jun 2025 01:40

Last Modified:

20 Jun 2025 01:40

URI:

https://repository.upnjatim.ac.id/id/eprint/38562