TCFimt: Temporal Counterfactual Forecasting from Individual Multiple Treatment Perspective

TL;DR

This paper introduces TCFimt, a novel framework for estimating causal effects of multiple treatments over time from individual data, addressing biases and interactions to improve prediction accuracy and treatment decision-making.

Contribution

The paper presents a comprehensive temporal counterfactual forecasting framework that employs adversarial training and contrastive learning to disentangle and accurately estimate multiple treatment effects.

Findings

Outperforms state-of-the-art methods in outcome prediction.

Effectively disentangles treatment effects and causal interactions.

Demonstrates applicability on real-world datasets from different fields.

Abstract

Determining causal effects of temporal multi-intervention assists decision-making. Restricted by time-varying bias, selection bias, and interactions of multiple interventions, the disentanglement and estimation of multiple treatment effects from individual temporal data is still rare. To tackle these challenges, we propose a comprehensive framework of temporal counterfactual forecasting from an individual multiple treatment perspective (TCFimt). TCFimt constructs adversarial tasks in a seq2seq framework to alleviate selection and time-varying bias and designs a contrastive learning-based block to decouple a mixed treatment effect into separated main treatment effects and causal interactions which further improves estimation accuracy. Through implementing experiments on two real-world datasets from distinct fields, the proposed method shows satisfactory performance in predicting future outcomes with specific treatments and in choosing optimal treatment type and timing than state-of-the-art methods.