基于决策树的手写数字的识别  

Handwritten Numeral Recognition Based On Decision Tree

作　　者：黄威 韦鹏[1] 

机构地区：[1]武汉大学电子信息学院,武汉430079

出　　处：《信息技术与信息化》2011年第6期21-23,73,共4页Information Technology and Informatization

摘　　要：手写数字的识别是模式识别及机器学习的一个重要应用,应用范围非常广泛。本文提出一种基于决策树算法的手写数字识别方法,该方法通过提取基于密度的特征,通过训练得到一个决策树分类模型,进而进行手写数字的识别。实验证明该方法能够快速有效的进行手写数字的识别。Handwritten numeral recognition is an application of pattern recognition and machine learning and has wide use in many fields. This paper proposed a method based on decision tree. The method can get a decision tree model for identification of handwritten numeral by extracting features based on： density. The experiments show that the proposed method can quickly and effectively recognize handwritten numeral.

关 键 词：手写数字识别 特征提取 模式识别 决策树 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]