KDD - arXivDaily 期刊&会议

2606.30992 2026-07-01 stat.ME cs.LG stat.AP 新提交

Hierarchical Clustering As a Novel Solution to the Notorious Multicollinearity Problem in Observational Causal Inference

层次聚类作为观测因果推断中多重共线性问题的新解决方案

Yufei Wu, Zhiying Gu, Alex Deng, Jacob Zhu, Linsha Chen

机构 * Airbnb, Inc.（Airbnb公司）

AI总结针对观测因果推断中多重共线性导致无法分离变量影响的问题，提出基于层次聚类聚合数据以缓解共线性的方法，并通过营销应用验证其有效性。

Comments Presented at the KDD 2023 Workshop on Causal Inference and Machine Learning in Practice, Long Beach, CA; also presented at the 2023 Joint Statistical Meetings

2606.30664 2026-07-01 stat.AP cs.AI cs.LG 新提交

Estimating the Effect of Timing on Coupon Effectiveness

估计时机对优惠券有效性的影响

Deddy Jobson

机构 * Mercari, Inc.（Mercari公司）

AI总结提出一个因果推断框架，利用自然随机对照试验估计在关键时机发送优惠券的效果，无需专用AB测试，并通过案例和公开数据集验证其有效性。

Comments 12 pages, 5 figures. Published in Proceedings of the 1st Workshop on End-End Customer Journey Optimization, co-located with KDD 2022, August 15, 2022, Washington, DC

2606.30999 2026-07-01 cs.LG econ.EM stat.AP stat.ME 新提交

Estimating Supply Incrementality in Two-sided Marketplaces: A Causal Machine Learning Approach

双边市场中供给增量的估计：一种因果机器学习方法

Yufei Wu, Daniel Schmierer, Dan Zylberglejd

机构 * Airbnb, Inc.（爱彼迎公司）

AI总结本文提出一种结合双重/去偏机器学习与层次贝叶斯框架的因果方法，利用地理空间相似性度量估计双边市场中新增供给对交易量的影响，并在Airbnb数据上验证了其合理性和强样本外性能。

Comments 5 pages, 3 figures. Accepted at the KDD 2025 Workshop on Causal Inference and Machine Learning in Practice (not presented)

2606.30932 2026-07-01 cs.LG stat.AP stat.ME 新提交

Personalizing Marketplace Policies with Competing Objectives and Constrained Experiments: Evidence from a Job Marketplace

具有竞争目标和受限实验的市场政策个性化：来自求职市场的证据

Yufei Wu, Zhen Yan

机构 * LinkedIn Corporation（领英公司）

AI总结针对双边市场政策个性化中跨方外部性和市场干扰问题，提出集成框架，通过混合排名模型和目标外推方法，在满足约束条件下提升目标指标。

Comments 10 pages, 6 figures. Accepted at ACM SIGKDD 2026 (Applied Data Science Track)

2606.30840 2026-07-01 cs.AI 新提交

Contrastive Reflection for Iterative Prompt Optimization

对比反思：迭代提示优化

Derek Koh, Jinghui Mo, Benjamin H. Le, Jiening Zhan, Baofen Zheng, Kevin Bevis, Nathaniel C. Owen, Lauren Elizabeth Charney, Wenqiong Liu, Jingwei Wu

机构 * LinkedIn（领英）

AI总结提出对比反思框架，通过对比失败与成功行为，迭代优化检索增强生成代理的提示，在HotpotQA上准确率从51.4%提升至60.4%。

Comments 6 pages, 1 figure. To appear at Agent4IR @ KDD 2026 (KDD 2026 Workshop on AI Agents for Information Retrieval)