Tensor flow小案例——04简易手写数字识别

最新推荐文章于 2022-05-20 15:44:05 发布

hxp1024

最新推荐文章于 2022-05-20 15:44:05 发布

阅读量299

点赞数 1

CC 4.0 BY-SA版权

分类专栏： Tensor Flow小案例

本文链接：https://blog.youkuaiyun.com/iv__vi/article/details/82899690

Tensor Flow小案例专栏收录该内容

6 篇文章

订阅专栏

本文介绍了一个使用TensorFlow实现的手写数字识别模型，通过构建神经网络并在MNIST数据集上进行训练，逐步提高了识别准确率，最终达到90%以上。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

# 定义神经网络中的隐藏层
def add_layer(inputs, size_inputs, size_outputs, act_fun = None):
    Weight = tf.Variable(tf.random_normal([size_inputs, size_outputs]))
    biases = tf.Variable(tf.zeros([1, size_outputs]) + 0.1)
    Wx_plus_b = tf.matmul(inputs, Weight) + biases
    if act_fun == None:
        output = Wx_plus_b
    else:
        output = act_fun(Wx_plus_b)
    return output

# 计算正确率
def compute_accuracy(v_xs, v_ys):
    global prediction
    y_pre = sess.run(prediction, feed_dict={xs: v_xs})
    correct_prediction = tf.equal(tf.argmax(y_pre, 1), tf.argmax(v_ys, 1))
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
    result = sess.run(accuracy, feed_dict={xs: v_xs, ys: v_ys})
    return result

# 引入数据
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)


# 定义输入变量
xs = tf.placeholder(tf.float32, [None, 784], name='x_in')
ys = tf.placeholder(tf.float32, [None, 10], name='y_in')

# 预测值
prediction = add_layer(xs, 784, 10, act_fun=tf.nn.softmax)

# 定义代价函数
cross_entropy = tf.reduce_mean(-tf.reduce_sum(ys * tf.log(prediction), reduction_indices=[1]))

# 定义训练方法：梯度下降；  指定学习率：0.001（一定不要太大，否则会NAN）；  训练目的：最小化loss
train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)


# 初始化全部变量
init = tf.global_variables_initializer()
sess = tf.Session()
sess.run(init)

# 迭代5,000次
for i in range(5000):
    batch_xs, batch_ys = mnist.train.next_batch(100)
    sess.run(train_step, feed_dict={xs: batch_xs, ys: batch_ys})
    # sess.run(train_step, feed_dict={xs: x_data, ys: y_data})
    # 输出正确率
    if i % 50 == 0:
        print(compute_accuracy(mnist.test.images, mnist.test.labels))

sess.close()

结果：

0.1072
0.6214
0.7308
0.7728
0.7939
0.8129
0.8296
0.8383
0.8468
0.8515
0.8545
0.858
0.8628
0.8616
0.868
0.8686
0.871
0.8717
0.8752
0.8734
0.8782
0.8806
0.8787
0.8795
0.8792
0.8763
0.8838
0.8799
0.8808
0.8849
0.8826
0.8829
0.8863
0.8887
0.8888
0.8878
0.8867
0.8888
0.8901
0.8918
0.8864
0.8889
0.8905
0.8922
0.8896
0.8903
0.8905
0.8934
0.8911
0.8919
0.8924
0.8922
0.8923
0.897
0.8961
0.897
0.8952
0.8966
0.8968
0.8953
0.8981
0.8984
0.8961
0.8998
0.8977
0.8987
0.8968
0.8986
0.8978
0.9001
0.9016
0.9001
0.9015
0.8989
0.9015
0.8988
0.8978
0.9023
0.9023
0.902
0.903
0.8986
0.8971
0.9024
0.9029
0.9047
0.9058
0.9045
0.9054
0.9
0.9041
0.9059
0.9038
0.9016
0.9035
0.9041
0.9076
0.9057
0.9036
0.9058