| GA改进的统计组合模型在个人信用评估中的应用 |
| |
| 引用本文： | 姜明辉,袁绪川.GA改进的统计组合模型在个人信用评估中的应用[J].哈尔滨工业大学学报(社会科学版),2007,9(1):97-100. |
| |
| 作者姓名： | 姜明辉 袁绪川 |
| |
| 作者单位： | 哈尔滨工业大学,管理学院,哈尔滨,150001 |
| |
| 基金项目： | 技术.政策.管理(TPM)国家哲学社会科学创新(HTCSR06T06) |
| |
| 摘 要： | 将分类精度作为遗传算法的适应度函数,利用遗传算法改进了基于线性回归和Logistic回归的统计组合预测模型,并将其应用于个人信用评估。结果表明,遗传算法改进后的组合预测模型在总分类精度和两类误判率方面都低于单一模型以及基于误差平方和最小的组合预测模型。利用遗传算法改进组合预测模型并用于个人信用评估是具有优势的。
|
| 关 键 词： | 个人信用评估 组合预测 遗传算法 |
| 文章编号： | 1009-1971(2007)01-0097-04 |
| 修稿时间： | 2006年9月10日 |
Application of Combining Forecast Model Based on Statistical Methods Optimized by GA in Personal Credit Scoring |
| |
| JIANG Ming-hui,YUAN Xu-chuan.Application of Combining Forecast Model Based on Statistical Methods Optimized by GA in Personal Credit Scoring[J].Journal of Harbin Institute of Technology(Social Sciences Edition),2007,9(1):97-100. |
| |
| Authors: | JIANG Ming-hui YUAN Xu-chuan |
| |
| Abstract: | Using predictive accuracy as the fitness function of genetic algorithm,this paper established a combining forecast model based on linear regression and logistic regression and used this combining forecast model in personal credit scoring.The results indicate that the combining model has a higher predictive accuracy and lower misclassification rates than each single model and the combining model based on the least sum of square error.The combining forecast model based on genetic algorithm has advantages in personal credit scoring. |
| |
| Keywords: | personal credit scoring combining forecast GA |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
