基于AlexNet模型的手写数字识别，用Tensorflow实现，可直接运行，准确率达到98 以上-Model based AlexNet handwritten digit recognition, implemented Tensorflow, can be run directly, the accuracy rate of more than 98

用卷积神经网络实现的手写数字识别（minst）,可直接运行，识别率较高。用Tensorflow实现-Handwritten digital recognition (minst) with convolution neural network can be run directly and the recognition rate is high. Implemented with Tensorflow

用简单神经网络(CNN)实现的手写数字识别(minst),可直接运行，用的是Tensorflow框架，通过该例程可以掌握Tensorflow基本用法-With a simple neural network (CNN) to achieve the handwritten numeral recognition (minst), can be run directly, using the Tensorflow framework, through the routine can master