基于信息增益的最优组合因子Fisher判别法

摘要

Linear discriminant classifier is an effective model analysis technology, among them with Fisher discriminant criterion, the most widely used, there are many kinds of improved linear extraction method. This paper introduces the infor-mation gain, to establish the optimal combination of the factors based on the information gain discriminant classifier, quickly chooses the optimal discriminant classifier combination factor. Experiments show that the rapid selected classifier eliminates the redundancy factor effectively, and has good classification accuracy.%线性判别分类器是一种有效的模式分析技术，其中以Fisher判别法准则应用最广，目前已有多种改进线性提取方法。引进信息增益，建立基于信息增益的最优组合因子判别分类器，实现最优组合因子判别分类器的优化。实验结果表明优化后的分类器有效剔除了冗余因子，具有良好的分类准确率。