Keras 功能 API：采用多个输入的拟合和测试模型答案

【问题标题】：Keras functional API: fitting and testing model that takes multiple inputsKeras 功能 API：采用多个输入的拟合和测试模型
【发布时间】：2018-12-23 02:54:50
【问题描述】：

我构建了一个具有 2 个分支的 Keras 模型，每个分支对相同的数据采用不同的特征表示。任务是将句子分为 6 个类别之一。

我已经测试了高达model.fit 的代码，它接收一个包含两个输入特征矩阵的列表作为X。一切正常。但是在预测时，当我将两个输入特征矩阵传递给测试数据时，就会产生错误。

代码如下：

X_train_feature1 = ... # shape: (2200, 100) each row a sentence and each column a feature
X_train_feature2 = ... # shape: (2200, 13) each row a sentence and each column a feature
y_train= ... # shape: (2200,6)


X_test_feature1 = ... # shape: (587, 100) each row a sentence and each column a feature
X_test_feature2 = ... # shape: (587, 13) each row a sentence and each column a feature
y_test= ... # shape: (587,6)

model= ... #creating a model with 2 branches, see the image below

model.fit([X_train_feature1, X_train_feature2],y_train,epochs=100, batch_size=10, verbose=2) #Model trains ok
model.predict([X_test_feature1, X_test_feature2],y_test,epochs=100, batch_size=10, verbose=2) #error here

模型如下所示：

错误是：

predictions = model.predict([X_test_feature1,X_test_feature2], y_test, verbose=2)
  File "/home/zz/Programs/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 1748, in predict
    verbose=verbose, steps=steps)
  File "/home/zz/Programs/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 1290, in _predict_loop
    batches = _make_batches(num_samples, batch_size)
  File "/home/zz/Programs/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 384, in _make_batches
    num_batches = int(np.ceil(size / float(batch_size)))
TypeError: only length-1 arrays can be converted to Python scalars

如果能帮助我理解错误以及如何修复它，我将不胜感激。

【问题讨论】：

标签： python machine-learning keras nlp classification

【解决方案1】：

predict method 仅将数据（即x）和batch_size（不必设置）作为输入。它不接受标签或时期作为输入。

如果你想预测类，那么你应该使用predict_classes 方法，它会为你提供预测的类标签（而不是predict 方法给出的概率）：

preds_prob = model.predict([X_test_feature1, X_test_feature2])
preds = model.predict_classes([X_test_feature1, X_test_feature2])

如果您想在测试数据上评估您的模型以找到损失和指标值，那么您应该使用evaluate 方法：

loss_metrics = model.evaluate([X_test_feature1, X_test_feature2], y_test)

【讨论】：