【新手发文】keras自定义了precision、recall指标，为啥这俩指标都是一样的数值？

weixin_45268706 2020-06-03 11:48:35

网上抄的指标代码，应该没问题，但是为啥运行中显示的这俩指标都是一样的？还定义了一个f1，也和他们一样？有大佬知道为什么吗？

def check_units(y_true, y_pred):

    if y_pred.shape[1] != 1:

      y_pred = y_pred[:,1:2]

      y_true = y_true[:,1:2]

    return y_true, y_pred



def precision(y_true, y_pred):

    y_true, y_pred = check_units(y_true, y_pred)

    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))

    predicted_positives = K.sum(K.round(K.clip(y_pred, 0, 1)))

    precision = true_positives / (predicted_positives + K.epsilon())

    return precision



def recall(y_true, y_pred):

    y_true, y_pred = check_units(y_true, y_pred)

    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))

    possible_positives = K.sum(K.round(K.clip(y_true, 0, 1)))

    recall = true_positives / (possible_positives + K.epsilon())

    return recall

下面是运行部分截图

  12/846 [..............................] - ETA: 6:37 - loss: 5.2509 - accuracy: 0.1624  - precision: 0.1624 - recall: 0.1624 - f1: 0.1624

 13/846 [..............................] - ETA: 6:23 - loss: 5.2181 - accuracy: 0.1641  - precision: 0.1641 - recall: 0.1641 - f1: 0.1641

 14/846 [..............................] - ETA: 6:16 - loss: 5.1733 - accuracy: 0.1666  - precision: 0.1666 - recall: 0.1666 - f1: 0.1666

 15/846 [..............................] - ETA: 6:05 - loss: 5.1398 - accuracy: 0.1686  - precision: 0.1686 - recall: 0.1686 - f1: 0.1686

 16/846 [..............................] - ETA: 5:55 - loss: 5.1052 - accuracy: 0.1703  - precision: 0.1703 - recall: 0.1703 - f1: 0.1703

 17/846 [..............................] - ETA: 5:46 - loss: 5.0709 - accuracy: 0.1722  - precision: 0.1722 - recall: 0.1722 - f1: 0.1722

...全文