基于深度学习的手写数字识别技术应用

Application of Handwritten Digit Recognition Technology Based on Deep Learning

本文基于深度学习技术搭建卷积神经网络(Convolutional Neural Networks-简称CNN),用MNIST数据集作为训练以及测试样本设计了一个手写数字识别技术应用系统。系统采用ReLu的激活函数,实验结果表明对手写数字的识别准确精度可以达到98.6%。

In this paper, we build a Convolutional Neural Networks(CNN) based on deep learning technology, and designs a handwritten digit recognition system. We use MNIST data set as our training and testing set. ReLu activation function is employed to improve recognition accuracy.Experimental results show that the accuracy of recognition of handwritten digits can be 98.6%.