Akbar, Moch. Januar (2025) Pemodelan Hybrid Transfer Function Dengan Neural Network Untuk Peramalan Outflow dan Inflow Uang Kartal pada Provinsi Bali. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5003211073-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5003211073-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy |
Abstract
Perkembangan ekonomi merupakan salah satu tujuan utama NKRI untuk salah satu upaya sesuai pada pembukaan UUD 1945 dalam Alinea keempat dan Pasal 33 UUD 1945. Salah satu indikator ekonomi adalah peredaran uang kartal yang mencerminkan pola inflow dan outflow terutama di provinsi Bali sebagai destinasi wisata favorit oleh masyarakat mancanegara pada negara Indonesia. Penelitian ini menggunakan model Hybrid dengan menggabungkan model fungsi transfer sebagai model linear dengan neural network sebagai model non-linear untuk meningkatkan akurasi hasil peramalan inflow dan outflow uang kartal pada provinsi Bali. Analisis dilakukan dengan mempertimbangkan efek variasi kalender dan variabel eksogen seperti inflasi. Berdasarkan penelitian sebelumnya, kombinasi model linear dan nonlinear serta penggunaan variabel eksogen terbukti memberikan hasil peramalan yang lebih akurat dibanding dengan hanya menggunakan model linear atau model nonlinear. Metode evaluasi yang digunakan meliputi Root Mean Square Error (RMSE), Mean Absolute Deviation (MAD) dan Akaike Information Criterion (AIC). Model terbaik untuk inflow diperoleh dari model Hybrid time series regression neural network, namun karena ketidaksignifikanan parameter secara parsial, model restricted digunakan sebagai alternatif. Meskipun peforma pada evaluasi dengan MAD sebesar 352,8 dan RMSE sebesar 445,64 untuk inflow pada outsample pada outsample sedikit lebih rendah. Peramalan inflow dan outflow pada model restricted Hybrid transfer function neural netwrok dengan nilai untuk testing atau outsample MAD sebesar 393,93 dan RMSE sebesar 469,97 pada outflow beserta MAD sebesar 366,9 dan RMSE sebesar 453,52 pada inflow. Model restricted menunjukkan bahwa variabel inflasi kurang relevan untuk peramalan inflow dan outflow
====================================================================================================================================
Preamble of the 1945 Constitution, specifically in the fourth paragraph and Article 33 of the 1945 Constitution. One of the economic indicators is the circulation of physical cash currency, which reflects the patterns of inflow and outflow, particularly in Bali Province, a favorite tourist destination for international visitors to Indonesia. This study employs a Hybrid model combining the transfer function model as a linear model and the neural network as a non-linear model to enhance the accuracy of forecasting the inflow and outflow of physical cash currency in Bali Province. The analysis incorporates the effects of calendar variations and exogenous variables such as inflation. Based on previous research, the combination of linear and nonlinear models, along with the inclusion of exogenous variables, has been shown to produce more accurate forecasts compared to using only linear or nonlinear models. The evaluation methods used include Root Mean Square Error (RMSE), Mean Absolute Deviation (MAD), and Akaike Information Criterion (AIC). The best model for inflow was obtained using the hybrid time series regression neural network model. However, due to the insignificance of some parameters in the partial tests, a restricted model was used as an alternative. Although its performance in the out-of-sample evaluation for inflow, with a MAD of 352.8 and an RMSE of 445.64, was slightly lower, the restricted hybrid transfer function neural network model showed acceptable results for both inflow and outflow forecasts. For the testing or out-of-sample data, the model produced a MAD of 393.93 and an RMSE of 469.97 for outflow, and a MAD of 366.9 and an RMSE of 453.52 for inflow. The restricted model also indicated that the inflation variable was not relevant for forecasting inflow and outflow.
Actions (login required)
 |
View Item |