N-gram and neural network language models differ in the way they calculate probability and train models.

• A.True
• B.False

Comment: *

Name: *

Email: *

Verification: *

During a training, the ReLU neuron cannot be activated on all training data. We call this the dead ReLu problem.

• A.False
• B.True

Comment: *

Name: *

Email: *

Verification: *

When using ModelArts automatic learning to build a predictive analysis project, after the model training is completed for the classification problem, which of the following options will be displayed in the evaluation results? (Multiple choice)

• A.Recall
• B.F1 value
• C.Precision
• D.Accuracy

Comment: *

Name: *

Email: *

Verification: *

In the NNLM model, what is the activation function used in the network output layer?

• A.Sigmoid
• B.ReLU
• C.Tanh
• D.Softmax

Comment: *

Name: *

Email: *

Verification: *

Forward propagation can complete the update of parameters.

• A.True
• B.False

Comment: *

Name: *

Email: *

Verification: *