Deep Learning-Driven Enterprise Sustainability Risk Prediction and Management

Abstract

The further development of global sustainable development objectives has led to the fact that the problem of sustainability risk management in enterprises has become a central area of contemporary corporate governance. Based on the prediction of corporate sustainability risks and their management, the present paper suggests a deep learning-based approach to enterprise sustainability risk prediction and management integrating multi-sector Environmental, Social, and Governance (ESG) data and a hybrid deep learning model based on Convolutional Neural Networks (CNN) as well as Long Short-Term Memory networks (LSTM). Experimental analysis on historic data of 1,200 companies across the world shows that both the training and validation demonstrated that the proposed solution approaches the risk prediction accuracy of 94.2%, an 18.5% increase over conventional machine learning methods with an F1-score of 0.923. The research results can be an intelligent way to manage the enterprise sustainability risk, it can help enterprise to identify and eliminate the possible sustainability risks as early as possible so that long term sustainable development of enterprise would be achieved.