keras用auc做metrics以及早停实例
我就废话不多说了,大家还是直接看代码吧~
importtensorflowastf fromsklearn.metricsimportroc_auc_score defauroc(y_true,y_pred): returntf.py_func(roc_auc_score,(y_true,y_pred),tf.double) #BuildModel... model.compile(loss='categorical_crossentropy',optimizer='adam',metrics=['accuracy',auroc])
完整例子:
defauc(y_true,y_pred): auc=tf.metrics.auc(y_true,y_pred)[1] K.get_session().run(tf.local_variables_initializer()) returnauc defcreate_model_nn(in_dim,layer_size=200): model=Sequential() model.add(Dense(layer_size,input_dim=in_dim,kernel_initializer='normal')) model.add(BatchNormalization()) model.add(Activation('relu')) model.add(Dropout(0.3)) foriinrange(2): model.add(Dense(layer_size)) model.add(BatchNormalization()) model.add(Activation('relu')) model.add(Dropout(0.3)) model.add(Dense(1,activation='sigmoid')) adam=optimizers.Adam(lr=0.01) model.compile(optimizer=adam,loss='binary_crossentropy',metrics=[auc]) returnmodel ####cvtrain folds=StratifiedKFold(n_splits=5,shuffle=False,random_state=15) oof=np.zeros(len(df_train)) predictions=np.zeros(len(df_test)) forfold_,(trn_idx,val_idx)inenumerate(folds.split(df_train.values,target2.values)): print("foldn°{}".format(fold_)) X_train=df_train.iloc[trn_idx][features] y_train=target2.iloc[trn_idx] X_valid=df_train.iloc[val_idx][features] y_valid=target2.iloc[val_idx] model_nn=create_model_nn(X_train.shape[1]) callback=EarlyStopping(monitor="val_auc",patience=50,verbose=0,mode='max') history=model_nn.fit(X_train,y_train,validation_data=(X_valid,y_valid),epochs=1000,batch_size=64,verbose=0,callbacks=[callback]) print('\nValidationMaxscore:{}'.format(np.max(history.history['val_auc']))) predictions+=model_nn.predict(df_test[features]).ravel()/folds.n_splits
补充知识:Keras可使用的评价函数
1:binary_accuracy(对二分类问题,计算在所有预测值上的平均正确率)
binary_accuracy(y_true,y_pred)
2:categorical_accuracy(对多分类问题,计算在所有预测值上的平均正确率)
categorical_accuracy(y_true,y_pred)
3:sparse_categorical_accuracy(与categorical_accuracy相同,在对稀疏的目标值预测时有用)
sparse_categorical_accuracy(y_true,y_pred)
4:top_k_categorical_accuracy(计算top-k正确率,当预测值的前k个值中存在目标类别即认为预测正确)
top_k_categorical_accuracy(y_true,y_pred,k=5)
5:sparse_top_k_categorical_accuracy(与top_k_categorical_accracy作用相同,但适用于稀疏情况)
sparse_top_k_categorical_accuracy(y_true,y_pred,k=5)
以上这篇keras用auc做metrics以及早停实例就是小编分享给大家的全部内容了,希望能给大家一个参考,也希望大家多多支持毛票票。