Interview Guides
You are working on an NLP classification problem where the text representation is the main design choice. Your team is deciding whether to use TF-IDF features or word embeddings before training a downstream model.
How would you choose between TF-IDF and word embeddings?