Scholarlink

<p>In the context of global sustainable development, Environmental, Social, and Governance (ESG) investment has emerged as a critical focus within the fields of finance and economics. This study aims to apply deep learning techniques to predict ESG ratings with enhanced accuracy and interpretability. We particularly examine the role of temporal information in prediction models and the importance of various factors influencing ESG assessments. The results demonstrate that Recurrent Neural Network models, particularly the GRU variant, outperform the other models, achieving best accuracy rate. Additionally, we employ interpretability tools to identify the key determinants driving ESG ratings. These findings not only provide valuable insights for investors but also offer actionable implications for corporate managers and policymakers aiming to enhance sustainability practices.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predicting ESG Ratings with Recurrent Neural Network Models: Evidence from China’s A-share Listed Companies

  • 期刊论文

  • 发表日期：

    2026年1月16日

查看全文
  • Shun Chen,
  • Jinhong Zhang,
  • Lei Ge

摘要

In the context of global sustainable development, Environmental, Social, and Governance (ESG) investment has emerged as a critical focus within the fields of finance and economics. This study aims to apply deep learning techniques to predict ESG ratings with enhanced accuracy and interpretability. We particularly examine the role of temporal information in prediction models and the importance of various factors influencing ESG assessments. The results demonstrate that Recurrent Neural Network models, particularly the GRU variant, outperform the other models, achieving best accuracy rate. Additionally, we employ interpretability tools to identify the key determinants driving ESG ratings. These findings not only provide valuable insights for investors but also offer actionable implications for corporate managers and policymakers aiming to enhance sustainability practices.