一种基于粒子群算法的多重决策树分类器融合方法

针对数据挖掘中的分类问题,根据多分类器融合的思想,提出一种基于粒子群优化算法的多重决策树分类器融合方法。先将概率度量水平的多重决策树进行线性组合,然后在融合算法中采用粒子群算法优化连接权值矩阵。并在UCI标准数据集上对模型进行了实验研究。结果表明该融合分类方法比单个决策树分类方法具有更高的分类精度。

For classification problems in data mining. Based on thought of multiple classifiers combination method, this paper proposes a combination classification method of multiple decision trees based on PSO Algorithm. In the proposed multiple classifiers combination method, multiple decision trees that adopt the method of probability measurement level output are combined. Then PSO algorithm isused for the optimization of connection weight matrix in combination algorithm. Further more, An experimental investigation is performed on the UCI datasets. Results of the experiments indicate that the proposed combination classification method has higher classification accuracy level than single decision tree.