Rahino, Wahyu Ageng (2026) Forecasting The Unemployment Rate Of Sidoarjo Regency Using Gated Recurrent Unit Method With Bayesian Hyperparameter Optimization. Undergraduate thesis, UPN VETERAN JAWA TIMUR.
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (Cover)
20081010145-Cover..pdf Download (1MB) |
|
Text (bab 1)
20081010145-bab1.pdf Download (207kB) |
|
Text (bab 2)
20081010145-bab2.pdf Restricted to Repository staff only until 17 June 2028. Download (459kB) |
|
Text (bab 3)
20081010145-bab3.pdf Restricted to Repository staff only until 17 June 2028. Download (420kB) |
|
Text (bab 4)
20081010145-bab4.pdf Restricted to Repository staff only until 17 June 2028. Download (4MB) |
|
Text (bab 5)
20081010145-bab5.pdf Download (226kB) |
|
Text (daftar pustaka)
20081010145-daftarpustaka.pdf Download (178kB) |
|
Text (lampiran)
20081010145-lampiran.pdf Restricted to Repository staff only until 17 June 2028. Download (307kB) |
Abstract
Unemployment is a crucial indicator that reflects the quality of regional development, including in Sidoarjo Regency, East Java. This study aims to develop a multivariate time series forecasting model to predict the Open Unemployment Rate (TPT) in Sidoarjo by integrating macroeconomic variables such as GDP Growth, Investment, Working-Age Population, and Human Development Index (HDI). This study uses annual data for the period 2005 to 2024 obtained from the Central Statistics Agency (BPS) of Sidoarjo Regency. The methodology used is the Gated Recurrent Unit (GRU) model, a type of Recurrent Neural Network (RNN) designed to capture non-linear patterns and long-term dependence on sequential data. To improve model performance, Bayesian Hyperparameter Optimization is applied to find the optimal configuration of key parameters such as the number of GRU units, window size, learning rate, dropout rate, number of epochs, and batch size. Data goes through the pre-processing stage in the form of Min-Max normalization and sequence formation using the sliding window technique. Model performance is evaluated using Root Mean Square Error (RMSE). The test results of 729 hyperparameter combinations showed that the best configuration was found with a combination of 16 GRU units, window size 2, learning rate 0.02, dropout rate 0.5, 100 epoch, and batch size 4. The multivariate GRU model in the best configuration produces an RMSE value of 0.067978 with a forecasting accuracy of 94.68%. These results indicate that the integration of GRU with Bayesian optimization provides an accurate method for forecasting the unemployment rate, as well as offering insights for data-driven policy formulation in Sidoarjo Regency.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | Q Science > QA Mathematics > QA76 Computer software Q Science > QA Mathematics > QA76.6 Computer Programming |
||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Informatics | ||||||||||||
| Depositing User: | Wahyu Ageng Rahino | ||||||||||||
| Date Deposited: | 18 Jun 2026 07:54 | ||||||||||||
| Last Modified: | 18 Jun 2026 08:17 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/54038 |
Actions (login required)
 |
View Item |