基于决策二叉树的多类支持向量机在大气质量评定中的应用

经典的支持向量机(Support Vector Machine，简称SVM)是针对二类分类的，在解决大气质量等级这种典型的多类分类问题时存在困难，本文提出了基于决策二叉树支持向量机的大气质量分类算法，将分类器分布在各个结点上，从而构成了多类支持向量机，减少了分类器数量和重复训练样本的数量。实验结果表明该方法能够正确地对大气质量进行识别。

THE APPLICATION TO ATMOSPHERE QUALITY ASSESSMENT BASED ON MULTI-CLASS SVM-DECISION BINARY TREE

The traditional Support Vector Machine s only deal with the binary classification.It has difficulty in solving the multi - class Atmosphere Quality Assessment problems.This paper presents an Atmosphere Q uality Assessment classification algorithm which is based on SVM-decision bina ry tree.This method distributes classifier to each nodes which constitutes multi -class SVM.It can reduce the number of SVM classifier and duplicating training samples.The experiment shows that this method is precised.