← 研踪 PaperTrailarXiv:2605.26288 · v2
Beyond Differences: Estimating and Tuning Ratio-Based Treatment Effects
首次提交 2026-05-26 03:24
版本更新 2026-11-17 01:01
雷达观测 2026-11-17 16:10
摘要
When treatment effects are naturally expressed as ratios -- as in medicine, pricing, and marketing -- the ratio-based CATE is the appropriate estimand. Yet the common estimators divide by an estimated baseline, impose a parametric form, or regress the outcome, where a better fit need not mean a better effect estimate. We introduce the Q-Learner, which writes as a ratio of two odds, reducing ratio-CATE estimation for binary outcomes to two propensity classification tasks. In benchmarks on synthetic and randomized real-world data, the Q-Learner proves especially suited to industry practice. Standard hyperparameter tuning on its log-loss suffices to make it rank well, get the level of the effects right and their spread more reliably than the alternatives, whereas, for instance, the common S-Learner needs a costly, effect-specific tuning to rank well, often best, but then misses the spread of the effects. If the propensity is known, a single classifier fitted on only the converters is enough. This not only speeds up development, but is often a necessity in practice, where data are frequently collected only for converters.
作者备注:29 pages (14 main text + 13 appendix), 6 figures, 11 tables. Code: https://github.com/michaelfuchs90/ratiobasedcate
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-17 01:01 · 最新观测版本
Beyond Differences: Estimating and Tuning Ratio-Based Treatment Effects
When treatment effects are naturally expressed as ratios -- as in medicine, pricing, and marketing -- the ratio-based CATE is the appropriate estimand. Yet the common estimators divide by an estimated baseline, impose a parametric form, or regress the outcome, where a better fit need not mean a better effect estimate. We introduce the Q-Learner, which writes as a ratio of two odds, reducing ratio-CATE estimation for binary outcomes to two propensity classification tasks. In benchmarks on synthetic and randomized real-world data, the Q-Learner proves especially suited to industry practice. Standard hyperparameter tuning on its log-loss suffices to make it rank well, get the level of the effects right and their spread more reliably than the alternatives, whereas, for instance, the common S-Learner needs a costly, effect-specific tuning to rank well, often best, but then misses the spread of the effects. If the propensity is known, a single classifier fitted on only the converters is enough. This not only speeds up development, but is often a necessity in practice, where data are frequently collected only for converters.
查看此版本解读 →查看此版本原文 ↗