Using financial and sociological data from Türkiye between 1999 and 2024, this work aims to develop an artificial intelligence (AI) augmented budget forecasting model. It also seeks to evaluate sustainability using the output of this model. This study assessed the efficacy of a number of methods, including Artificial Neural Networks (ANN), Random Forest (RF), XGBoost (XGB), and Deep Learning (DL). The study employed SHapley Additive Explanations (SHAP), a method associated with explainable artificial intelligence (XAI), to explain the model's output. The findings indicate that the RF model is more effective at predicting borrowing needs, even while the XGB model offers the most accurate forecasts for overall revenues and expenses. The SHAP analysis emphasizes how important variables like loan amounts, debt levels, and demographic size are. The findings emphasize how crucial it is to combine explainable budget forecasts in Türkiye with AI techniques.
İnce et al. (Mon,) studied this question.