期刊文献+

基于遗传算法的属性约简 被引量:6

Genetic Algorithm Based Features Reduction
下载PDF
导出
摘要 属性约简是知识发现的重要步骤,但从属性集中选择最优子集属于NP-hard问题。文章提出的遗传属性抽取算法,以属性的可分性度量为偏置,并引入禁忌表搜索策略,降低了搜索空间;采用退火选择来保持种群的个体多样性,防止未成熟收敛;算法内置的分类器采用人工神经网络,并提出了基于有监督聚类的人工神经网络分类算法,有效地降低了人工神经网络分类器的训练时间。实验分析表明,算法能够从高维数、大数据集合中有效降低数据维数,提高数据的分类准确性。 Features Reduction is a crucial process in knowledge discovery, however to select an optimal subset from feature set is a NP-Hard problem. A neural network based classification module and GA-based property abstraction algorithm are designed and implemented, which takes the separability measurement of feature as the offset to decrease the searching space of GA, as well as combines the tabu-table searching to avoid repeated searching. In addition, anneal se- lection is introduced to maintain the diversity of populations, and prevent immaturity convergence. An artificial neural network based classification module is designed, and a supervisory clustering method is proposed to reduce its training time. The experimental result shows that this mechanism can reduce the data dimension and improve the accuracy of data classification.
出处 《微电子学与计算机》 CSCD 北大核心 2006年第7期150-153,共4页 Microelectronics & Computer
基金 国家"863"计划项目(2003AA001048)
关键词 遗传算法 属性约简 人工神经网络 Genetic algorithm, Features reduction, Artificial neural network
  • 引文网络
  • 相关文献

参考文献8

  • 1Weston J,Mukherjee S,Chapelle O,et al.Feature selection for SVMs[J].Advances in Neural Information Processing Systems,2000:668~674
  • 2Liu H,Setiono R.Incremental feature selection[J].Applied Intelligence.1998,9(3):217~230
  • 3Yang J,Honavar V.Feature subset selection using a genetic algorithm[J].IEEE Intelligent Systems.1998,13:44~49
  • 4Yang J,Parekh,Honavar V.DistAl:an inter-pattern distance-based constructive learning algorithm[R".Tech.Rep.ISU-CS-TR 97-05,Iowa State University,USA,1997
  • 5Tümer M B,Demir M C.A genetic approach to data dimensionality reduction using a special initial population[A] international Work-Conference on the Interplay between natural and artificial computation[C].Las Palmas,Canary Islands,Spain 2005:310~316
  • 6Pernkopf F,O'Leary P.Feature selection for classification using genetic algorithms with a novel encoding[A].The 9th international conference on computer analysis of images and patterns[C].Springer-Verlag,London,UK,2001:161~168
  • 7Liu H,Motoda H.Feature selection for knowledge discovery and data mining[M].Boston:Kluwer Academic Publishers,1998
  • 8Richeldi M,Lanzi P.Performing effective feature selection by investigating the deep structure of the data[A].The second international conference of knowledge discovery and data mining[C].AAAI Press,Menlo Park,California.1996:379~383

同被引文献49

引证文献6

二级引证文献28

;
使用帮助 返回顶部