tf学习（三）—— MNIST手写识别 上

MNIST-NameError: name ‘input_data’ is not defined解决办法

其实这是由于导入工具库后没有使用正确别名的原因，只要加入as input_data即可。
应改成如下代码：
import tensorflow.examples.tutorials.mnist.input_data as input_data

mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)

了解MNIST存储情况

分为训练数据与测试数据

train_images=mnist.train.images
train_labels = mnist.train.labels
test_images = mnist.test.images

test_labels = mnist.test.labels

print("train_images_shape:", train_images.shape)
print("train_labels_shape:", train_labels.shape)
print("test_images_shape:", test_images.shape)

print("test_labels_shape:", test_labels.shape)

import tensorflow.examples.tutorials.mnist.input_data as input_data

mnist=input_data.read_data_sets("MNIST_data/",one_hot=True)

结果

train_images_shape: (55000, 784)
train_labels_shape: (55000, 10)
test_images_shape: (10000, 784)
test_labels_shape: (10000, 10)

print("train_images:", train_images[1])

结果

train_images中28*28的图片像素值介于0-1之间

print("train_images:", train_images[0])

print("train_images:", train_labels[1])

结果

train_images: [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]

print("test_labels:", test_labels[0])

结果

test_labels: [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]