How do you handle the cold-start problem?

medium

Answer

$New user or item with no interactions.$
$Solutions: (1) Content-based features (user profile, item text/image \to embed) \to hybrid recsys.$
$(2) Popularity-based fallback.$
$(3) Bandits: explore new items early (Thompson sampling on Beta posteriors).$
$(4) Meta-learning to warm-start embeddings.$
$(5) Ask onboarding questions.$
$Every real production recsys hybridizes CF + content to handle cold-start.$

Check yourself — multiple choice

Nothing to do
Content-based hybrid + popularity fallback + bandits (Thompson) for exploration + meta-learning warm-start + onboarding; production always hybridizes
Random
Only CF

Cold-start: content hybrid + popularity + bandits + meta-learning.

#applications

214 interview questions in this topic.