###
DOI:
Journal of Software:2004.15(10):1479-1484

基于因果发现的神经网络集成方法
凌锦江,周志华
(南京大学,计算机软件新技术国家重点实验室,江苏,南京,210093)
Causal Discovery Based Neural Network Ensemble Method
LING Jin-Jiang,ZHOU Zhi-Hua
()
Abstract
Chart / table
Reference
Similar Articles
Article :Browse 3435   Download 3698
Received:August 04, 2003    Revised:June 10, 2004
> 中文摘要: 现有的神经网络集成方法主要通过扰动训练数据来产生精确且差异度较大的个体网络,从而获得较强的泛化能力.利用因果发现技术,在取样结果中找出类别属性的祖先属性,然后使用仅包含这些属性的数据生成个体网络,从而有效地将扰动训练数据与扰动输入属性结合起来,以产生精确度高且差异度大的个体.实验结果表明,该方法的泛化能力与当前一些流行的集成方法相当或更好.
Abstract:Current neural network ensemble methods usually generate accurate and diverse component networks by disturbing the training data, and therefore achieve strong generalization ability. In this paper, causal discovery is employed to discover the ancestor attributes of the class attribute on the results of the sampling process. Then, component neural networks are trained on the samples with only the ancestor attributes being used as inputs. Thus, the mechanism of disturbing the training data and the input attribute is combined to help generate accurate and diverse component networks. Experiments show that the generalization ability of the proposed method is better than or comparable to that of the ensembles generated by some prevailing methods.
文章编号:     中图分类号:    文献标志码:
基金项目:Supported by the National Natural Science Foundation of China under Grant No.60273033(国家自然科学基金);the National Outstanding Youth Foundation of China under Grant No.60325207(国家杰出青年科学基金);the Natural Science Foundation of Jiangsu Province of China under Grant No. BK2004001 (江苏省自然科学基金);the Fork Ying Tung Education Foundation under Grant No.91067 (霍英东基金);the Excellent Young Teachers Program of the MOE of China(国家教育部优秀青年教师资助计划) Supported by the National Natural Science Foundation of China under Grant No.60273033(国家自然科学基金);the National Outstanding Youth Foundation of China under Grant No.60325207(国家杰出青年科学基金);the Natural Science Foundation of Jiangsu Province of China under Grant No. BK2004001 (江苏省自然科学基金);the Fork Ying Tung Education Foundation under Grant No.91067 (霍英东基金);the Excellent Young Teachers Program of the MOE of China(国家教育部优秀青年教师资助计划)
Foundation items:
Reference text:

凌锦江,周志华.基于因果发现的神经网络集成方法.软件学报,2004,15(10):1479-1484

LING Jin-Jiang,ZHOU Zhi-Hua.Causal Discovery Based Neural Network Ensemble Method.Journal of Software,2004,15(10):1479-1484