|
|||||
|
|
| Logistic模型选择中三种交叉验证策略的比较 | |
| 引用本文： | 家会臣,靳竹萱,李济洪.Logistic模型选择中三种交叉验证策略的比较[J].太原师范学院学报(自然科学版),2012,11(1):87-90. |
| 作者姓名： | 家会臣 靳竹萱 李济洪 |
| 作者单位： | 1. 山西大学数学科学学院,山西太原,030006 2. 北京大学数学科学学院,北京,100871 3. 山西大学计算中心,山西太原,030006 |
| 基金项目： | 国家自然科学基金(60873128) |
| 摘 要： | 在模型选择中,常用5折、10折交叉验证方法.文章给出一种基于3×2交叉验证的模型选择方法,并通过模拟实验证明了在Logistic模型中,3×2交叉验证要比5折和10折交叉验证选到真模型的概率更大. |
| 关 键 词： | 交叉验证 模型选择 logistic回归 R软件 |
The Comparison of Three Cross-Validation Strategy for Logistic Model Identification |
|
| Jia Huichen,Jin Zhuxuan,Li Jihong.The Comparison of Three Cross-Validation Strategy for Logistic Model Identification[J].Journal of Taiyuan Normal University:Natural Science Edition,2012,11(1):87-90. | |
| Authors: | Jia Huichen Jin Zhuxuan Li Jihong |
| Institution: | 1.School of Mathematical Sciences,Shanxi University,Taiyuan 030006; 2.School of Mathematical Sciences,Peking University,Beijing 100871; 3.Computer Center of Shanxi University,Taiyuan 030006,China) |
| Abstract: | The 5 fold and 10 fold cross validation methods are often employed in the model identification task.The 3×2 cross validation methods are introduced to identify the true model.The experimental results demonstrated that in the logistic regression framework,the probability of true model identification based on 3×2 cross validation is higher than that of 5 fold and 10 fold cross validation methods. |
| Keywords: | cross-validation model selection Logistic regression R software |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |