Bird
0
0

Why do RNNs for text classification often use a sigmoid activation in the output layer for binary classification?

hard📝 Conceptual Q10 of 15
NLP - Sequence Models for NLP
Why do RNNs for text classification often use a sigmoid activation in the output layer for binary classification?
ASigmoid outputs a probability between 0 and 1 for the positive class
BSigmoid speeds up training by normalizing inputs
CSigmoid prevents overfitting by limiting output range
DSigmoid converts text into numerical vectors
Step-by-Step Solution
Solution:

  1. Step 1: Understand sigmoid output role

    Sigmoid activation outputs values between 0 and 1, interpretable as probabilities for binary classes.

  2. Step 2: Eliminate incorrect reasons

    Sigmoid does not speed training, prevent overfitting, or convert text to vectors.

  3. Final Answer:

    Sigmoid outputs a probability between 0 and 1 for the positive class -> Option A

  4. Quick Check:

    Sigmoid = Probability output for binary class [OK]
Quick Trick: Sigmoid outputs probability for binary classification [OK]
Common Mistakes:
MISTAKES
  • Thinking sigmoid normalizes inputs
  • Believing sigmoid prevents overfitting
  • Confusing sigmoid with embedding

Want More Practice?

15+ quiz questions · All difficulty levels · Free

Free Signup - Practice All Questions
More NLP Quizzes
Sentiment Analysis Advanced - Sentiment with context (sarcasm, negation) - Quiz 15hard Sequence Models for NLP - LSTM for text - Quiz 8hard Sequence Models for NLP - GRU for text - Quiz 2easy Text Generation - Evaluating generated text (BLEU, ROUGE) - Quiz 11easy Text Generation - Evaluating generated text (BLEU, ROUGE) - Quiz 14medium Text Similarity and Search - Jaccard similarity - Quiz 13medium Topic Modeling - Topic coherence evaluation - Quiz 10hard Topic Modeling - Why topic modeling discovers themes - Quiz 14medium Word Embeddings - Word2Vec (CBOW and Skip-gram) - Quiz 3easy Word Embeddings - Word2Vec (CBOW and Skip-gram) - Quiz 15hard