ML
ML Interview Roadmap
← Study plan·MLOps

Day 65 of 133

Online experimentation for ML + DSA 1-D DP finish

A/B for models vs UI; shadow / canary / interleaving / off-policy.

DSA · NeetCode 1-D DP

  • Longest Increasing SubsequenceDSA · 1-D DP

    Interview questions to prep

    1. Walk through DP O(n²). Then walk through patience-sorting O(n log n) with binary search.
    2. What's the DIFFERENCE between LIS length and the actual subsequence — how do you reconstruct?
  • Partition Equal Subset SumDSA · 1-D DP

    Interview questions to prep

    1. State the DP: define the state, the transition, and the base case explicitly.
    2. Top-down (memoized recursion) vs bottom-up (tabulation) — which is more natural here, and why?
    3. Can you space-optimize from O(n) to O(1)? Show the rolling-window trick.

MLOps · Online experimentation

  • A/B testing for models: lift, guardrails, rampMLOpsEugene Yan

    Interview questions to prep

    1. How is A/B testing a model different from A/B testing a UI feature?
    2. What guardrail metrics do you track on every model launch?
  • Canary, shadow, interleavingMLOpsChip Huyen

    Interview questions to prep

    1. When does interleaving beat A/B for ranking systems?
    2. What's the smallest canary slice that's still statistically meaningful for a ranking model?
  • Counterfactual & off-policy evaluationMLOpsEugene Yan

    Interview questions to prep

    1. What is off-policy evaluation, and where does it show up in recommender systems?
    2. Compare IPS vs doubly-robust off-policy estimators — when does each fall apart?

References & further reading