相关实验视频
Updated: Jan 11, 2026

05:37
An R-Based Landscape Validation of a Competing Risk Model
Published on: September 16, 2022
2.5K
使用GPT-4和改进的BERT融合模型进行财务文本分析和信用风险评估
1School of Economics and Management, Shanghai Zhongqiao Vocational and Technical University, Shanghai City, China.
PloS one
|November 18, 2025
概括
本研究引入了使用生成预训练变压器-4 (GPT-4) 和BERT进行混合模型,以提高金融文本中的信用风险识别. 该方法提高了准确性和可解释性,用于更好的金融技术和自然语言处理应用程序.
科学领域:
- 金融技术 (金融科技)
- 自然语言处理 (NLP) 是一种自然语言处理.
- 计算金融是指计算金融.
背景情况:
- 传统的信用风险评估与非结构化财务数据作斗争.
- 局限性包括不良的上下文推理,行业术语识别和隐含风险表达.
- 现有的方法在复杂的金融文本中缺乏适应性和解释性.
研究的目的:
- 开发一种先进的混合模型,以便在非结构化金融文本中更好地识别信用风险.
- 克服传统方法在分析长文本和特定行业术语方面的局限性.
- 提高信用风险评估的上下文理解,行业适应性和解释性.
主要方法:
- 一个混合模型,将生成预训练变压器-4 (GPT-4) 结合起来,用于语义理解和从变压器 (BERT) 获得双向编码器表示,用于特征提取.
- 整合了金融词典增强模块和命名实体识别 (NER) 组件.
- 利用GPT-4进行基于提示的生成,以提取隐藏的风险信息,并使用双模型语义融合机制,以重视多层次风险评估的注意力.
主要成果:
- 与基线模型相比,拟议的混合模型表现出优越的性能.
- 在信用风险评估中实现了更高的准确性,适应性和解释性.
- 通过对公共金融数据集和现实世界年度报告的实验验验证的有效性.
结论:
- 混合GPT-4和BERT模型显著提高了在非结构化财务文本中的信用风险识别.
- 该方法为金融文本分析提供了一个强大的解决方案,解决了以前方法的关键局限性.
- 这项研究对金融科技和NLP的进步具有理论和实际意义.
相关概念视频
Mathematical Modeling: Problem Solving
231
Mathematical modeling transforms real-world scenarios into mathematical expressions, allowing for structured problem-solving and analysis. This process involves defining the situation, assigning variables to measurable quantities, selecting an appropriate model, and solving the resulting equation. Such models are invaluable in finance, providing precise methods to evaluate investments, loans, and repayment structures.A widely used example is the calculation of fixed monthly payments on a loan,...
231
Survival Tree
374
Survival trees are a non-parametric method used in survival analysis to model the relationship between a set of covariates and the time until an event of interest occurs, often referred to as the "time-to-event" or "survival time." This method is particularly useful when dealing with censored data, where the event has not occurred for some individuals by the end of the study period, or when the exact time of the event is unknown.
Building a Survival Tree
Constructing a...
Building a Survival Tree
Constructing a...
374
