Predicting the impact of treatments over time with uncertainty aware neural differential equations

TL;DR

This paper introduces CF-ODE, a neural differential equation model that predicts treatment effects over time with uncertainty quantification, improving accuracy and reliability in observational data scenarios.

Contribution

The paper presents a novel Neural ODE-based method, CF-ODE, that models treatment impacts continuously over time with uncertainty estimates, addressing confounding issues.

Findings

CF-ODE outperforms existing methods in prediction accuracy.

CF-ODE provides reliable uncertainty estimates.

The method is effective across multiple longitudinal datasets.

Abstract

Predicting the impact of treatments from observational data only still represents a majorchallenge despite recent significant advances in time series modeling. Treatment assignments are usually correlated with the predictors of the response, resulting in a lack of data support for counterfactual predictions and therefore in poor quality estimates. Developments in causal inference have lead to methods addressing this confounding by requiring a minimum level of overlap. However,overlap is difficult to assess and usually notsatisfied in practice. In this work, we propose Counterfactual ODE (CF-ODE), a novel method to predict the impact of treatments continuously over time using Neural Ordinary Differential Equations equipped with uncertainty estimates. This allows to specifically assess which treatment outcomes can be reliably predicted. We demonstrate over several longitudinal data sets that CF-ODE provides more accurate predictions and more reliable uncertainty estimates than previously available methods.