Predictive analytics framework for decision support in ethiopia’s foreign exchange market
| dc.contributor.advisor | Kuribachew Gizaw (PhD) | |
| dc.contributor.author | Hundera Belema | |
| dc.date.accessioned | 2026-11-06T04:57:28Z | |
| dc.date.available | 2026-11-06T04:57:28Z | |
| dc.date.issued | 2026-11-03 | |
| dc.description.abstract | This research designs and tests a predictive analytics framework that can be used to support decision-making in Ethiopia's foreign exchange market. The research utilises data mining techniques along with conventional econometrics to better estimate and identify foreign exchange determinants under realistic institutional constraints. Daily USD/ETB and monthly macroeconomic indicators from August 2024 to December 2025 are used and xgboost, ARIMA, and Prophet models are compared. Among the models xgboost achieves the highest accuracy (MAPE 2.62%, R² 0.6329), marginally outperforming ARIMA (2.71%), while Prophet fails. Based on SHAP analysis inflation is identified as the dominant driver of foreign exchange with 55% of feature importance followed by exports and imports. The final decision support system is built to be modular, adaptive and scalable. Therefore this study showcases the possibility of using data mining in Ethiopia's foreign exchange market improvement | |
| dc.identifier.uri | https://etd.aau.edu.et/handle/123456789/9359 | |
| dc.language.iso | en_US | |
| dc.publisher | ADDIS ABABA UNIVERSITY | |
| dc.subject | Predictive analytics | |
| dc.subject | foreign exchange | |
| dc.subject | Ethiopia | |
| dc.subject | XGBoost | |
| dc.subject | SHAP | |
| dc.subject | decision support system | |
| dc.subject | USD/ETB | |
| dc.subject | inflation | |
| dc.subject | machine learnin | |
| dc.title | Predictive analytics framework for decision support in ethiopia’s foreign exchange market | |
| dc.type | Thesis |