期刊文献+

基于构造超平面的两阶段决策树算法的研究 被引量:1

Two-stage Decision Tree Algorithm Based on Hyperplane Construction
下载PDF
导出
摘要 如何在测试节点里构造一个恰当的分割超平面是构造决策树的关键,与单变量决策树不同,多变量(倾斜)决策树可以找到与特征轴不垂直的超平面。本文将从几何学角度说明构造测试节点的过程,提出了一种两阶段决策树的算法。 How to construct an appropriate partitioning hyperplane in test node is the key to construct a decision tree.Different from decision tree with a single variable,the multi-variable(tilted) decision tree can find a hyperplane which is not perpendicular to the characteristic shaft.This paper will explain the process of constructing the test node and propose a two-stage decision tree algorithm.
出处 《价值工程》 2011年第8期168-168,共1页 Value Engineering
关键词 超平面 两阶段 决策树 hyperplane two stages decision tree
  • 相关文献

参考文献4

  • 1Su,X.G.,Tsai,C.-L,& Wang,C.(2009).Tree-structured model diagnostics for linear regression.Math Learn 74:111-131.
  • 2Vens,C.Struyf,J.,Schietgat,L,Dzeroski,S.,& Blockeel,H.(2008).Decision trees for hierarchical multi-label classification.Mach Learn,73:185-214.
  • 3Quinlan,J.R(1979).Discovering rules by induction from large collection of examples.In D.Michie,editor.
  • 4Quinlan J R.(1993).CA.5:Programs for Machine Learning[M].San Mateo,CA:Morgan Kaufmann.

同被引文献2

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部