Explainable AI-Integrated Fiscal Impact Prediction Framework for Budget Planning

This paper gives an Explainable AI-Integrated Fiscal Impact Prediction Framework that can help in the betterment of budget planning by making accurate, scalable, and explanatory predictions. The framework uses Robust Scaler as a preprocessing tool, which is useful in dealing with outliers and training a reliable model. The Principal Component Analysis is used in feature selection to reduce the dimensionality without losing some important information thus enhancing high computing efficiency. It is an effective gradient boosting classifier, which uses the XGBoost in order to extract more complicated trends in the fiscal data and this results in high accuracy and strength. This framework is deployed through TensorFlow and TensorFlow Extendedwhich facilitates easy deployment, scaling, and continuous updates of models. Also, explainable AI methods like SHAP and LIME are implemented, which offer insight into the decision-making process of the model which is essential in policy and budgetary decisions. The new methodology has provided a new, flexible approach to fiscal forecasting that would guarantee predictive accuracy and model understandability in the context of applicability.