A NARX-NN optimization algorithm for forecasting inflation during a potential recession period using longitudinal data

Restu Arisanti, Yahma Nurhasanah, Sri Winarni

Abstract


The global economy is facing the potential of a recession in 2023, and inflation is one of the factors that could trigger it.  This study focuses on forecasting inflation using longitudinal data and the NARX NN method, which combines Generalized Linear Mixed Model (GLMM) and Neural Network (NN) approaches. The accuracy of the NARX NN method's prediction results will be measured using evaluation values such as RMSE and MAE. The aim is to provide insight into the credibility of the potential recession in the next few years. The major findings from this study are as follows: 1) The best performing Feed Forward Neural Network (FFNN) model is FFNN (7-15-5), which was applied to all exogenous variables data and achieved RMSE values of X1 = 5.158, X2 = 7.377, X3 = 4.054, X4 = 0.456, X5= 5.130 and MAE values of X1 = 3.533, X2 = 4.667, X3 = 2.522, X4 = 0.216, X5 = 4.101; 2) The NARX NN series parallel model was utilized to forecast the USD/IDR exchange rate and its relationship with the exogenous variables, resulting in the best model of NARX NN (9-5-1). This model was applied to all inflation rate data and produced RMSE of 3,375 and MAE of 2,552; 3) Based on the forecasting results for the next 5 years (2022-2026), inflation rate is expected to experience an upward trend, indicating the possibility of an economic recession.

Full Text: PDF

Published: 2023-05-01

How to Cite this Article:

Restu Arisanti, Yahma Nurhasanah, Sri Winarni, A NARX-NN optimization algorithm for forecasting inflation during a potential recession period using longitudinal data, Commun. Math. Biol. Neurosci., 2023 (2023), Article ID 43

Copyright © 2023 Restu Arisanti, Yahma Nurhasanah, Sri Winarni. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Commun. Math. Biol. Neurosci.

ISSN 2052-2541

Editorial Office: [email protected]

 

Copyright ©2024 CMBN