On Inductive Biases for Heterogeneous Treatment Effect Estimation
arXiv:2106.03765
Abstract
We investigate how to exploit structural similarities of an individual's potential outcomes (POs) under different treatments to obtain better estimates of conditional average treatment effects in finite samples. Especially when it is unknown whether a treatment has an effect at all, it is natural to hypothesize that the POs are similar - yet, some existing strategies for treatment effect estimation employ regularization schemes that implicitly encourage heterogeneity even when it does not exist and fail to fully make use of shared structure. In this paper, we investigate and compare three end-to-end learning strategies to overcome this problem - based on regularization, reparametrization and a flexible multi-task architecture - each encoding inductive bias favoring shared behavior across POs. To build understanding of their relative strengths, we implement all strategies using neural networks and conduct a wide range of semi-synthetic experiments. We observe that all three approaches can lead to substantial improvements upon numerous baselines and gain insight into performance differences across various experimental settings.
arXiv commentTo Appear in the Proceedings of the 35th Conference on Neural Information Processing Systems (NeurIPS 2021)
References in corpus (8)
- Meta-learners for Estimating Heterogeneous Treatment Effects using Machine Learning
- Estimating treatment effect heterogeneity in randomized program evaluation
- Discussion on "Bayesian Regression Tree Models for Causal Inference: Regularization, Confounding, and Heterogeneous Effects" by Hahn, Murray and Carvalho
- Causal Effect Inference with Deep Latent-Variable Models
- The Benefit of Multitask Representation Learning
- Deep Counterfactual Networks with Propensity-Dropout
- Hypothesis Transfer Learning via Transformation Functions
- Learning Weighted Representations for Generalization Across Designs