RETRACTED: Advancing enterprise risk management with deep learning: A predictive approach using the...

Area of Science:

• Artificial Intelligence
• Financial Risk Management
• Machine Learning

Background:

• Traditional enterprise risk management faces limitations in complex markets.
• Diverse risk events require advanced detection and prediction methods.
• Sustainable development hinges on effective risk mitigation strategies.

Purpose of the Study:

• To introduce a deep learning-based risk management model.
• To enhance the prediction and detection of enterprise risk events.
• To address limitations of traditional risk management approaches.

Main Methods:

• Utilized a novel XGBoost-CNN-BiLSTM framework.
• Combined XGBoost for structured data, CNN for feature extraction, and BiLSTM for time series analysis.
• Validated the model on multiple datasets, including S&P 500 historical data.

Main Results:

• Achieved significant advantages in accuracy, recall, F1 score, and AUC.
• Demonstrated high precision (93.84%) and recall (95.75%) on the S&P 500 dataset.
• Verified the model's robustness and superiority in risk event prediction.

Conclusions:

• The XGBoost-CNN-BiLSTM model provides a more reliable method for enterprise risk management.
• Deep learning offers significant potential for advancing risk management practices.
• The study offers valuable insights for applying AI in financial risk assessment.