Y

LLMs and NLP · Intermediate · 5 h

Text as data

Tokenise, vectorise and classify text with TF-IDF and with embeddings, and compare them honestly.

Why employers care: A strong baseline is the first thing reviewers look for in any language task. It is often good enough to ship.

In the roadmaps for: Data Scientist, AI/ML Engineer

Lessons, in order

  1. Text classification with TF-IDF and embeddingsDetect spam in 5,572 real SMS messages two ways, and find out why the older method wins here.3 min

Next shelf: Calling LLMs in production code