python - 基于tensorflow对MNIST中的手写数字进行分类，但是对训练集与测试集的像素归一化后，训练时打印显示训练集的准确地极其低！这是为什么？

Question

我想基于tensorflow对MNIST中的手写数字进行分类，但是对训练集与测试集的像素归一化后，训练时打印显示训练集的准确地极其低！请大家帮我分析下原因，代码如下：

# 导入各种包
import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt
from tensorflow.examples.tutorials.mnist import input_data
import os
import pickle

# 超参数设置
numClasses = 10
inputSize = 784
batch_size = 64
learning_rate = 0.05

# 下载数据集
mnist = input_data.read_data_sets('original_data/', one_hot=True)

train_img = mnist.train.images
train_label = mnist.train.labels
test_img = mnist.test.images
test_label = mnist.test.labels
train_img /= 255.0
test_img /= 255.0


X = tf.compat.v1.placeholder(tf.float32, shape=[None, inputSize])
y = tf.compat.v1.placeholder(tf.float32, shape=[None, numClasses])
W = tf.Variable(tf.random_normal([inputSize, numClasses], stddev=0.1))
B = tf.Variable(tf.constant(0.1), [numClasses])
y_pred = tf.nn.softmax(tf.matmul(X, W) + B)

loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=y_pred)) + 0.01 * tf.nn.l2_loss(W)
opt = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss)

correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_pred, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))

saver = tf.train.Saver()
multiclass_parameters = {}
# 运行
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())

    # 开始训练 外循环控制轮数，内循环训练
    for epoch in range(20):
        total_batch = int(len(train_img) / batch_size)  # 总批次=训练的样本总数/批次大小

        for batch in range(total_batch):  # 一个批次的训练
            # 从训练集获取到一个批次的数据
            batch_input = train_img[batch * batch_size: (batch + 1) * batch_size]
            batch_label = train_label[batch * batch_size: (batch + 1) * batch_size]

            _, trainingLoss = sess.run([opt, loss], feed_dict={X: batch_input, y: batch_label})

        train_acc = sess.run(accuracy, feed_dict={X: train_img, y: train_label})
        print("Epoch %d Training Accuracy %g" % (epoch + 1, train_acc))

打印结果如下图所示：

显示准确率非常的低，请各位大佬帮忙分析下原因，感谢！

Answer 1

# 这行代码改一下：
# y_pred = tf.nn.softmax(tf.matmul(X, W) + B)
# 改为
y_pred = tf.matmul(X, W) + B

# 其他代码保持不变

这些修改看看：

# ...（导入包和设置超参数的代码部分保持不变）

# 下载数据集
mnist = input_data.read_data_sets('original_data/', one_hot=True)

train_img = mnist.train.images
train_label = mnist.train.labels
test_img = mnist.test.images
test_label = mnist.test.labels
train_img /= 255.0
test_img /= 255.0

X = tf.compat.v1.placeholder(tf.float32, shape=[None, inputSize])
y = tf.compat.v1.placeholder(tf.float32, shape=[None, numClasses])
W = tf.Variable(tf.random_normal([inputSize, numClasses], stddev=0.1))
B = tf.Variable(tf.constant(0.1), [numClasses])
y_pred = tf.matmul(X, W) + B

loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=y_pred)) + 0.01 * tf.nn.l2_loss(W)
opt = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss)

correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(tf.nn.softmax(y_pred), 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))

saver = tf.train.Saver()
multiclass_parameters = {}

# ...（运行代码部分保持不变）