EasyDeepLearn

How do you handle cold-start users / items in production?

medium

Answer

  • (1) Content-based fallback: use features (demographics, item metadata) via a separate model.
  • (2) Global popular items.
  • (3) Simple heuristics (category-popular).
  • (4) Bandits / Thompson sampling to explore fast.
  • (5) Two-tower model that generalizes via features.
  • (6) Hybrid: gradually blend into personalized as history accumulates.
  • Cold-start is not a modeling failure — it's a system design pattern (product should still function).
Check yourself — multiple choice
  • No solution
  • Content-based fallback + global popular + heuristics + bandits + two-tower feature generalization + gradual blend; system pattern (product still functions)
  • Random
  • Wait for data

Cold-start: content fallback + popular + bandits + two-tower + blend.

#deployment#features

Practise MLOps & Data Quality

215 interview questions in this topic.

Related questions