Sklearn 与 Tensorflow 进行分类

在Skikit-learn中提供了一些函数来计算分类指标：准确率和召回率。

>>> from sklearn.metrics import precision_score, recall_score
>>> precision_score(y_train_5, y_pred) # == 4344 / (4344 + 1307)
0.76871350203503808
>>> recall_score(y_train_5, y_train_pred) # == 4344 / (4344 + 1077)
0.79136690647482011

F1值为准确率与召回率的平均。

>>> from sklearn.metrics import f1_score
>>> f1_score(y_train_5, y_pred)
0.78468208092485547