Active Learning for Optimal Intervention Design in Causal Models

TL;DR

This paper introduces a causal active learning method that efficiently identifies optimal interventions in complex systems by using a Bayesian approach and a novel acquisition function, demonstrated on synthetic and biological data.

Contribution

It develops a causally informed active learning strategy with a closed-form acquisition function, providing theoretical guarantees for optimal intervention design in causal models.

Findings

Outperforms existing methods in intervention selection efficiency.

Proven theoretical bounds and consistency for linear causal models.

Successfully applied to biological data for cell state transition interventions.

Abstract

Sequential experimental design to discover interventions that achieve a desired outcome is a key problem in various domains including science, engineering and public policy. When the space of possible interventions is large, making an exhaustive search infeasible, experimental design strategies are needed. In this context, encoding the causal relationships between the variables, and thus the effect of interventions on the system, is critical for identifying desirable interventions more efficiently. Here, we develop a causal active learning strategy to identify interventions that are optimal, as measured by the discrepancy between the post-interventional mean of the distribution and a desired target mean. The approach employs a Bayesian update for the causal model and prioritizes interventions using a carefully designed, causally informed acquisition function. This acquisition function is evaluated in closed form, allowing for fast optimization. The resulting algorithms are theoretically grounded with information-theoretic bounds and provable consistency results for linear causal models with known causal graph. We apply our approach to both synthetic data and single-cell transcriptomic data from Perturb-CITE-seq experiments to identify optimal perturbations that induce a specific cell state transition. The causally informed acquisition function generally outperforms existing criteria allowing for optimal intervention design with fewer but carefully selected samples.