Prediksi Laju Inflasi Kota Semarang Berdasarkan Fluktuasi Harga Komoditas Volatile Food Menggunakan Arsitektur Multivariat Long Short-Term Memory

Nastiti, Rahmah Lidya (2026) Prediksi Laju Inflasi Kota Semarang Berdasarkan Fluktuasi Harga Komoditas Volatile Food Menggunakan Arsitektur Multivariat Long Short-Term Memory. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 22083010102_cover.pdf Download (711kB)
	Text (Bab 1) 22083010102-bab1.pdf Download (186kB)
	Text (Bab 2) 22083010102-bab2.pdf Restricted to Repository staff only until 7 September 2029. Download (389kB)
	Text (Bab 3) 22083010102-bab3.pdf Restricted to Repository staff only until 7 September 2029. Download (297kB)
	Text (Bab 4) 22083010102-bab4.pdf Restricted to Repository staff only until 7 September 2029. Download (1MB)
	Text (Bab 5) 22083010102-bab5.pdf Download (14kB)
	Text (Daftar Pustaka) 22083010102-daftarpustaka.pdf Download (94kB)
	Text (Lampiran) 22083010102-lampiran.pdf Restricted to Repository staff only Download (178kB)

Abstract

Inflation is a key indicator in maintaining a region’s macroeconomic stability because it affects people’s purchasing power and economic policymaking. In Semarang, inflation is influenced by price fluctuations in volatile food commodities—particularly rice, chicken, and red chili peppers—which exhibit dynamic patterns and inter-variable relationships that must be considered in predictive modeling. Therefore, a predictive model capable of simultaneously utilizing temporal patterns and information from multiple variables is required. This study aims to develop a predictive model for Semarang’s inflation rate using Multivariate Long Short-Term Memory (LSTM) and to compare its performance with that of a Vector Autoregressive (VAR) model. The research methodology includes data preprocessing, VAR modeling, and Multivariate LSTM modeling using three architectures: 1-layer, 2-layer, and 3-layer LSTM. Evaluation was conducted using MAE, RMSE, and MAPE. The results show that the 3-layer Multivariate LSTM model delivered the best performance with an MAE of 0.3721, an RMSE of 0.5543, and a MAPE of 1.3990%. This model also produced lower error values compared to the VAR model, which had an MAE of 0.4380, an RMSE of 0.5652, and a MAPE of 4.8148%. Thus, the 3-layer Multivariate LSTM was selected as the model with the best performance in this study.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Hindrayani, Kartika Maulida	NIDN0009099205	kartika.maulida.ds@upnjatim.ac.id
Thesis advisor	Wara, Shindi Sella May	NUPTK1850774675230252	shindi.shella.fasilkom@upnjatim.ac.id

Subjects:

H Social Sciences > H Social Sciences (General)
H Social Sciences > HA Statistics
H Social Sciences > HB Economic Theory
Q Science > QA Mathematics
Q Science > QA Mathematics > QA76.6 Computer Programming

Divisions:

Faculty of Computer Science > Departemen of Data Science

Depositing User:

Mahasiswa Rahmah Lidya Nastiti

Date Deposited:

07 Sep 2026 02:19

Last Modified:

07 Sep 2026 02:19

URI:

https://repository.upnjatim.ac.id/id/eprint/59967