Scholarlink

<p>Along with the expansion of debt markets worldwide in recent years, the default crisis caused by credit debt increasingly impacts the financial system and people’s livelihoods. Banks require an effective and efficient credit evaluation system to assess corporate creditworthiness. This paper presents a methodology for rating the reputation of companies using Convolutional Neural Networks with co-optimized structural parameters. The method conducts big data processing and co-optimizes structural parameters using transaction invoice information and historical reputation rating data from previous companies. A feature selection technique based on the random forest method is introduced to improve predictive fitting accuracy. Finally, the model enhances robustness through decision integration. Experimental results show that the model performs well in corporate reputation rating, ultimately achieving 91 percent accuracy.</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

Research on Enterprise Credit Rating Method based on Structural Parameter Co-optimization Convolutional Neural Network

  • 期刊论文

  • 发表日期：

    2025年10月8日

查看全文
  • Xuanyun Liu

摘要

Along with the expansion of debt markets worldwide in recent years, the default crisis caused by credit debt increasingly impacts the financial system and people’s livelihoods. Banks require an effective and efficient credit evaluation system to assess corporate creditworthiness. This paper presents a methodology for rating the reputation of companies using Convolutional Neural Networks with co-optimized structural parameters. The method conducts big data processing and co-optimizes structural parameters using transaction invoice information and historical reputation rating data from previous companies. A feature selection technique based on the random forest method is introduced to improve predictive fitting accuracy. Finally, the model enhances robustness through decision integration. Experimental results show that the model performs well in corporate reputation rating, ultimately achieving 91 percent accuracy.