Just Trial Once: Ongoing Causal Validation of Machine Learning Models

TL;DR

This paper proposes a method to evaluate the causal impact of updated machine learning models using data from prior randomized controlled trials, reducing the need for costly new experiments.

Contribution

It introduces conditions under which the causal effects of new ML model versions can be estimated or bounded from existing RCT data, considering model updates and user trust.

Findings

Provides bounds for causal impact of new ML models without new RCTs

Recommends trial design strategies to better assess future model updates

Offers practical guidelines to save resources in model deployment

Abstract

Machine learning (ML) models are increasingly used as decision-support tools in high-risk domains. Evaluating the causal impact of deploying such models can be done with a randomized controlled trial (RCT) that randomizes users to ML vs. control groups and assesses the effect on relevant outcomes. However, ML models are inevitably updated over time, and we often lack evidence for the causal impact of these updates. While the causal effect could be repeatedly validated with ongoing RCTs, such experiments are expensive and time-consuming to run. In this work, we present an alternative solution: using only data from a prior RCT, we give conditions under which the causal impact of a new ML model can be precisely bounded or estimated, even if it was not included in the RCT. Our assumptions incorporate two realistic constraints: ML predictions are often deterministic, and their impacts depend on user trust in the model. Based on our analysis, we give recommendations for trial designs that maximize our ability to assess future versions of an ML model. Our hope is that our trial design recommendations will save practitioners time and resources while allowing for quicker deployments of updates to ML models.