Precision@K: Fraction of the top K results that are relevant - how accurate are the recommendations?
Recall@K: Fraction of all relevant items that made it into the top K - how complete is retrieval? This is the core metric for the retrieval stage.
NDCG: Gain with graded relevance and position discount - relevant items ranked higher score more. Normalized by the ideal DCG to 0-1. It handles good results must also be in the right position.
Positions: retrieval uses Recall@K, ranking uses NDCG, and Precision@K when display slots are limited.
Get asked follow-ups live, then receive a scored report — like a real MLE interview loop.
Start AI mock interview