跳到正文
返回论文列表
econ.EM提交于 已译

实用 DML 与 AI 学习表示

Pragmatic DML with AI-Learned Representations

Andres Aradillas Fernandez · Victor Chernozhukov · Carlos Cinelli · Sven Klaassen · Whitney Newey · Martin Spindler · et al.

中文摘要

文本、图像等丰富对象越来越多地通过 AI 学习表示作为控制变量进入因果分析。对一大类估计量而言,用表示替代原始对象会使目标偏移,偏移量为两个误差的内积:一个来自结果回归,一个来自平衡权重。因此偏差受二者规模乘积的约束。由此可得三点结论。表示固定时,常规交叉拟合去偏机器学习(DML)区间对表示特定目标有效;当近似偏差相对抽样误差可忽略时,该区间对因果目标也有效。若该偏差不可忽略,敏感性分析可界定因果目标并为端点提供置信界。当嵌入在样本上调优时,三种方法可保持有效推断:固定菜单微调、联合损失调优和双重表示选择。一个多模态需求应用展示了该工作流程。

关键要点

  1. 01用 AI 学习表示替代原始对象作控制变量,会使因果目标偏移为结果回归误差与平衡权重误差的内积。
  2. 02表示固定时,交叉拟合 DML 区间对表示特定目标有效;近似偏差可忽略时对因果目标也有效。
  3. 03偏差不可忽略时,敏感性分析可界定因果目标并为端点给出置信界。
  4. 04嵌入在样本上调优时,固定菜单微调、联合损失调优和双重表示选择可保持有效推断。
  5. 05多模态需求应用展示了该工作流程。

解读

尚无解读。

原始英文摘要

arXiv:2610.01935v3 Announce Type: replace-cross Abstract: Text, images, and other rich objects increasingly enter causal analysis through AI-learned representations used as controls. For a broad class of estimands, replacing the raw object by its representation shifts the target by the inner product of two errors: one in the outcome regression and one in the balancing weight. The bias is therefore bounded by the product of their sizes. Three consequences follow. With a fixed representation, the usual cross-fitted debiased machine learning (DML) interval is valid for a representation-specific target, and for the causal target when the approximation bias is negligible relative to sampling error. When this bias may not be negligible, a sensitivity analysis bounds the causal target and supplies confidence bounds for the endpoints. When the embedding is tuned on the sample, three approaches preserve valid inference: fixed-menu fine-tuning, joint-loss tuning, and double representation selection. A multimodal demand application illustrates the workflow.

同方向论文 · econ.EM

查看全部 →