RealCause: Realistic Causal Inference Benchmarking

TL;DR

RealCause introduces a realistic causal inference benchmark using generative models that provides ground-truth causal effects, enabling more accurate evaluation of estimators on data resembling real-world scenarios.

Contribution

The paper presents a new benchmark for causal inference that combines ground-truth causal effects with realistic data generation, facilitating better estimator evaluation.

Findings

Over 1500 causal estimators evaluated.

Hyperparameter tuning via predictive metrics is justified.

Benchmark closely mimics real data characteristics.

Abstract

There are many different causal effect estimators in causal inference. However, it is unclear how to choose between these estimators because there is no ground-truth for causal effects. A commonly used option is to simulate synthetic data, where the ground-truth is known. However, the best causal estimators on synthetic data are unlikely to be the best causal estimators on real data. An ideal benchmark for causal estimators would both (a) yield ground-truth values of the causal effects and (b) be representative of real data. Using flexible generative models, we provide a benchmark that both yields ground-truth and is realistic. Using this benchmark, we evaluate over 1500 different causal estimators and provide evidence that it is rational to choose hyperparameters for causal estimators using predictive metrics.