Abstract
The currency crisis that hit Turkey in 1994, 2001, and 2018 caused the depreciation of the lira value and other negative impacts, especially on the economic sector. Efforts to overcome these losses by developing a model that can identify future currency crises. The aim of this research is to build an early detection model for the currency crisis in Turkey based on macroeconomic indicators using a multilayer perceptron backpropagation model through SGD, ADAM, NADAM, and AdaBound optimization. The independent variables used are 11 Turkish macroeconomic indicators for the period January 1990 to December 2022, while the response variable is a perfect signal value. The perfect signal value can be determined by the Financial Pressure Index (FPI). The analysis results show that the best model is the multilayer perceptron backpropagation model with NADAM optimization. Model testing with NADAM optimization on data validation obtained an accuracy of 97.29% and 93.33% on test data. The detection results show that from January 2023 to December 2024, there will be no currency crisis in Turkey.
Keywords:
Article Details
- Year: 2025
- Volume: 46
- Issue: 3
- Pages: 265 - 284
- Accepted: 30.08.2025
-
DOI:
https://doi.org/10.5281/zenodo.17416384 -
Full Text (PDF)
3 Views 1 Downloads
How to Cite?
Sugiyanto , Isnandar Slamet, Etik Zukhronah, Alvina Aulia Rahma, Ihsan Fathoni Amri (2025). Implementation of Backpropagation Artificial Neural Network Algorithm for Turkey Currency Crisis Detection Based on Financial Pressure Index. Journal of Economic Cooperation and Development, 46(3), 265-284. https://doi.org/10.5281/zenodo.17416384
