基于关键度度量的决策树算法改进及其在铁路运输中的应用

采用决策树方法对客票数据及行车安全数据进行分析时,发现在训练数据集的主类类属分布不平衡的情况下,无法对数据量占相对少数的小类属数据进行描述,究其原因在于现有决策树方法的节点类标号标示采用的是"多数表决"的方式。为此,本文提出关键度度量的概念,用于改进决策树的叶节点竞争机制,旨在改进决策树方法在解决弱势类属在数据建模分析中公平获得"发言权"的问题。算例表明,这种改进较好地解决了偏类数据集的数据分析问题。

Improved Algorithm of Decision Tree Based on Key Decision Factor and Its Applications in Railway Transportation

According to the analyses on applications of decision tree induction in railways,it is found that there is limitation in analyzing the imbalance-distribution main classes because "major voting" is selected as its leaf node measure.This paper presents the measure of using the key decision factor to improve the leaf node label measure of the decision tree,which aims at solving the unfair competition among the main classes between the majorities and the minorities for labeling the leaf nodes.This algorithm adapts to proceeding with the huge imbalance ticket datasets and extracts a kind of instructive rules that collect the advantages in both prediction and statistics,therefore it is suitable for supporting multi-level requirements of the decision-makers for predictive analysis.