Understanding Recurrent Neural Networks

What is the purpose of concatenating the outputs of the forward and backward RNNs?

• To form a combined representation

• To discard unnecessary information

• To simplify the model

• To increase computational cost

What are some applications of Bi-RNNs?

• Sentiment Analysis

• Image Classification

• Data Compression

• Time Series Forecasting

What is one advantage of Bidirectional RNNs?

• Enhanced Contextual Understanding

• Lower computational cost

• Simpler architecture

• Faster training times

What is a drawback of Bidirectional RNNs?

• Increased Computational Complexity

• Better accuracy

• More efficient memory usage

• Faster predictions

What types of Bi-RNNs are mentioned?

• LSTM Bi-RNNs

• GRU Bi-RNNs

• CNN Bi-RNNs

• SVM Bi-RNNs

What metrics are used to evaluate traditional language modeling?

• Accuracy

• F1-Score

• Perplexity

• Cross entropy

What is the perplexity metric in NLP?

• A way to capture the degree of uncertainty a model has in predicting text.

• A measure of the average uncertainty in predicting words.

• A metric used to evaluate the performance of language models.

How is perplexity computed over a sentence W?

• PP(W) = 1/Pnorm(W)

• PP(W) = P(W)^(1/n)

• PP(W) = Pnorm(W)^(1/n)

What does cross-entropy measure?

• How well the predicted distribution aligns with the true distribution.

• The average uncertainty in predicting words.

• The performance of language models.

What is Bits-per-Character (BPC)?

• A metric used to evaluate the performance of language models.

• A measure of the average number of bits required to represent a character in a given text.

• A method to compute the perplexity of a language model.