Machine learning h2o如何返回mcc值_Machine Learning_H2o

Machine learning h2o如何返回mcc值

machine-learning

Machine learning h2o如何返回mcc值,machine-learning,h2o,Machine Learning,H2o,我创建了随机forrest模型。mcc值是两个值的列表。为什么? mRF=H2ORandomForestEstimator(nfolds=10,keep_cross_validation_models = True,seed=12345,model_id="RF0",ntrees=1000) mRF.train(x=column_use, y=target, training_frame=train,validation_frame=valid) print("MCC valid",mRF.mo

我创建了随机forrest模型。mcc值是两个值的列表。为什么?

mRF=H2ORandomForestEstimator(nfolds=10,keep_cross_validation_models = True,seed=12345,model_id="RF0",ntrees=1000)
mRF.train(x=column_use, y=target, training_frame=train,validation_frame=valid)
print("MCC valid",mRF.model_performance(valid).mcc())
MCC valid [[0.35743618321170406, 0.21239407659849494]]

如果不设置参数threshold，它将返回一个数组，其中包含一个最优阈值的结果-[[optimal threshold，mcc]]。您可以设置自己的阈值数组-[threshold1，threshold2，…]，它将返回[[threshold1，mcc1]，[threshold2，mcc2]，…]