What is the difference between sparse and dense feature representations in machine learning?

Machine Learning Hard

Machine Learning — Hard

What is the difference between sparse and dense feature representations in machine learning?

Key points

  • Sparse features have few non-zero entries
  • Dense features capture richer distributed information
  • Sparse features are common in one-hot encoding and TF-IDF
  • Dense features are common in word embeddings and PCA output

Ready to go further?

Related questions