Exploring Artificial Intelligence Models for Early Warning Systems with Systemic Risk Analysis in Finance

This study examines early warnings for financial system risks using machine learning, with the goal of offering valuable insights and approaches to risk management in the finance industry. The study employs a systematic methodology to develop an AI-driven early warning system for systemic risk analysis in finance using the “Risk Analytics in Banking and Financial Services” dataset. Two machine learning models, CatBoost Classifier and Hist Gradient Boosting Classifier, were trained and evaluated according to ROC curve, f1-score, re-call, accuracy, and precision. The CatBoost Classifier achieved an accura-cy94.35<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>, precision99.93%, and recall88.74%, while the Hist Gradient Boosting Classifier outperformed with 95.05% accuracy, 99.92% precision, and 90.15% recall. Both models demonstrated high performance, with the Hist Gradient Boosting Classifier providing slightly better results compared to existing Deep FM and XGBoost accuracy, indicating its superior capability for accurate systemic risk prediction. The empirical findings demonstrate that this model provides favourable prediction effects on Financial system risk warning.