This paper introduces new methods for evaluating ad ranking models, focusing on counterfactual evaluation. It proposes using Inverse Propensity Scoring (IPS) and Doubly Robust (DR) estimators to assess model performance more accurately than traditional approaches. The goal is to provide a reliable way to determine if a new ranking model is superior to the current production model before deployment. AI
IMPACT Provides advanced evaluation techniques for ranking models, potentially improving ad targeting and performance.
RANK_REASON The cluster contains a research paper detailing new methods for model evaluation. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
