keras第一课：Cifar分类

from keras.datasets import cifar10  #keras自带mnist/cifar10/cifar100等数据集的演示功能
import numpy as np
#np.random.seed(10)  #随机数

数据准备

(x_img_train,y_label_train),(x_img_test,y_label_test)=cifar10.load_data()   

x_img_train_normalize = x_img_train.astype('float32') / 255.0  #数据类型转换，然后归一化
x_img_test_normalize = x_img_test.astype('float32') / 255.0

from keras.utils import np_utils
y_label_train_OneHot = np_utils.to_categorical(y_label_train)  #OneHot 编码
y_label_test_OneHot = np_utils.to_categorical(y_label_test)

建立模型

from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Conv2D, MaxPooling2D, ZeroPadding2D

model = Sequential()

#卷积层

model.add(Conv2D(filters=32,kernel_size=(3, 3),input_shape=(32, 32,3),activation='relu', padding='same'))
model.add(Dropout(0.3))
model.add(MaxPooling2D(pool_size=(2, 2)))

#卷积层

model.add(Conv2D(filters=64, kernel_size=(3, 3),activation='relu', padding='same'))
model.add(Dropout(0.3))
model.add(MaxPooling2D(pool_size=(2, 2)))

#节点数据展开

model.add(Flatten())
model.add(Dropout(0.3))

#全连接层，隐藏层

model.add(Dense(1024, activation='relu'))
model.add(Dropout(0.3))

#全连接层，输出层

model.add(Dense(10, activation='softmax'))

训练模型

model.compile(loss='categorical_crossentropy',optimizer='adam', metrics=['accuracy'])

train_history=model.fit(x_img_train_normalize, y_label_train_OneHot,validation_split=0.2,epochs=10, batch_size=128, verbose=1)

 

评估模型的准确率

scores = model.evaluate(x_img_test_normalize, y_label_test_OneHot, verbose=0)

进行预测

prediction=model.predict_classes(x_img_test_normalize)

 

简单体会：keras很好用，一目了然。但是里面的参数怎么选择？这个还需要深入学习。