6.keras-基于CNN网络的Mnist数据集分类

Posted 2020-12-09 wigginess

keras-基于CNN网络的Mnist数据集分类

1.数据的载入和预处理

import numpy as np
from keras.datasets import mnist
from keras.utils import np_utils
from keras.models import Sequential
from keras.layers import *
from keras.optimizers import SGD,Adam
from keras.regularizers import l2
from keras.utils.vis_utils import plot_model
from matplotlib import pyplot as plt

import os

import tensorflow as tf

# 载入数据
(x_train,y_train),(x_test,y_test) = mnist.load_data()

# 预处理
# 将（60000，28，28）转化为（-1,28,28,1），最后1是图片深度

x_train = x_train.reshape(-1,28,28,1)/255.0
x_test= x_test.reshape(-1,28,28,1)/255.0
# 将输出转化为one_hot编码
y_train = np_utils.to_categorical(y_train,num_classes=10)
y_test = np_utils.to_categorical(y_test,num_classes=10)

2.创建网络打印训练数据# 创建网络

model = Sequential([
   
    # 创建卷积层提取特征，对卷积核进行正则化
    Conv2D(input_shape=(28,28,1),filters=32,kernel_size=5,strides=1,padding=‘same‘,activation=‘relu‘,kernel_regularizer=l2(0.01)), 
　　 # 池化层，对特征的筛选
　　 MaxPool2D(pool_size=(2,2),strides=2,padding=‘same‘), Flatten(),
　　
　　 Dense(units=128,input_dim=784,bias_initializer=‘one‘,activation=‘tanh‘),

    Dropout(0.2),
    Dense(units=10,bias_initializer=‘one‘,activation=‘softmax‘)
])

# 编译
# 自定义优化器
sgd = SGD(lr=0.1)
adma = Adam(lr=0.001)
# 运用交叉熵
model.compile(optimizer=adma,
              loss=‘categorical_crossentropy‘,
              # 得到训练过程中的准确率
              metrics=[‘accuracy‘])

model.fit(x_train,y_train,batch_size=32,epochs=10,validation_split=0.2)

# 评估模型
loss,acc = model.evaluate(x_test,y_test,)
print(‘
test loss‘,loss)
print(‘test acc‘,acc)

out:

Epoch 1/10

32/48000 [..............................] - ETA: 10:18 - loss: 2.7563 - acc: 0.1562
96/48000 [..............................] - ETA: 3:53 - loss: 2.6141 - acc: 0.1354

......

......

Epoch 10/10

45952/48000 [===========================>..] - ETA: 0s - loss: 0.0664 - acc: 0.9905
47616/48000 [============================>.] - ETA: 0s - loss: 0.0663 - acc: 0.9908
48000/48000 [==============================] - 2s 37us/step - loss: 0.0663 - acc: 0.9910 - val_loss: 0.0149 - val_acc: 0.9884

 

32/10000 [..............................] - ETA: 4s
3360/10000 [=========>....................] - ETA: 0s
5824/10000 [================>.............] - ETA: 0s
8512/10000 [========================>.....] - ETA: 0s
10000/10000 [==============================] - 0s 20us/step

test loss 0.015059704356454312
test acc 0.988