Recurrent Neural Networks Quiz

What you'll be asked

1. What is the primary purpose of using a Recurrent Neural Network (RNN) compared to a standard feedforward neural network?
2. Which of the following scenarios is most appropriate for a Recurrent Neural Network architecture?
3. What is the primary challenge associated with the vanishing gradient problem in standard RNNs?
4. How do Long Short-Term Memory (LSTM) units address the limitations of vanilla RNNs?
5. If you are training an RNN and notice that the gradients are exploding (i.e., becoming extremely large), which technique is the most standard practical remedy?

Answers and explanations come after you submit.