基于深度学习的手写数字分类问题研究  

Research on the Classification of Handwriting Number Based on Deep Learning

作　　者：宋志坚[1] 余锐[2] 

机构地区：[1]重庆交通大学信息科学与工程学院,重庆400074 [2]重庆交通大学交通运输学院,重庆400074

出　　处：《重庆工商大学学报（自然科学版）》2015年第8期49-53,共5页Journal of Chongqing Technology and Business University:Natural Science Edition

基　　金：重庆市自然科学基金计划项目(CSTC2011jj A0893)

摘　　要：手写体数字因其书写风格差异大、上下文无关及识别准确度要求高等原因导致其识别难度大,针对手写体数字识别的特点及要求,使用深度学习算法进行分类,通过对样本的训练完成手写体数字的识别,同时与SVM算法及BP神经网络分类效果进行对比;实验结果表明深度学习在识别手写体数字时具有更高的准确率。It is difficult to recognize handwritingand recognition requires high accuracy. According tonumber, because writing styles are different, context is freethe characteristics and requirements of handwriting numberrecognition, this paper uses the Deep Learning algorithm for classification, based on the sample of training torecognize the handwriting number. At the same time, the experiment compared the classification effect with that of SVM algorithm and BP neuralnetwork. The experimental results show that the Deep Learning of handwriting number in recognition has a higheraccuracy.

关 键 词：手写体数字识别 深度学习 样本训练 

分 类 号：O224[理学—运筹学与控制论]