keras基本训练及注意事项

keras训练步骤

from keras.layers import Dense  #导入层
from keras.models import Sequential  

model = Sequential()
model.add(Dense(1, input_shape = (1, )))  #input_shape = (特征的维度,)
model.compile(loss = "mse", optimizer =  "sgd")

model.fit(X_train, y_train, epochs = 20, batch_size = 50) #可以加验证集

model.evaluate(X_test, y_test) #在验证集上的准确率
model.predict() #

## history = model.fit() 输出训练的准确率

keras交叉熵与独热编码
使用softmax激活函数进行多分类时，需要将标签转化成独热编码，然后使用交叉熵损失函数进行训练

from keras.utils.np_utils import to_categorical
feature, label = create_data()
X_train, X_test, y_train, y_test = train_test_split(feature, label, test_size = 0.2)
y_train = to_categorical(y_train) #转化成独热编码
y_test = to_categorical(y_test)

model = Sequential()
model.add(Dense(2,activation = "softmax", input_shape = (4, ) ))

#使用交叉熵作为损失函数
model.compile(optimizer = "rmsprop",loss = "categorical_crossentropy", metrics = ["accuracy"]) 

model.fit(X_train, y_train, epochs = 20)