目录前言···························1第一章绪论·······················21.1研究背景及意义································································21.2深度学习的发展现状··························································31.3字符识别发展及研究··························································41.4本文的主要组织结构··························································6第二章神经网络基础介绍·······················································82.1引言···············································································82.2神经网络介绍···································································82.2.1神经元基本简介······························································82.2.2前向传播算法（Forwardpropagation）·································92.2.3反向传播算法与梯度下降算法··········································112.3手写字符数据集简介························································132.4本章小结········································································14第三章基于卷积神经网络的手写字符识别································153.1引言··············································································153.2卷积··············································································153.3权值共享········································································163.4池化··············································································163.5LetNet识别手写字符·························································173.5.1使用LeNet进行训练·······················································173.5.2神经网络模型改进·························································183.5.3改进模型实验结果·························································193.5.4识别错误原因分析·························································203.5.5对影响收敛速度因素的实验测试····································...
