Dynamic Causal Bayesian Optimization

Abstract

This paper studies the problem of performing a sequence of optimal interventions in a causal dynamical system where both the target variable of interest and the inputs evolve over time. This problem arises in a variety of domains e.g. system biology and operational research. Dynamic Causal Bayesian Optimization (DCBO) brings together ideas from sequential decision making, causal inference and Gaussian process (GP) emulation. DCBO is useful in scenarios where all causal effects in a graph are changing over time. At every time step DCBO identifies a local optimal intervention by integrating both observational and past interventional data collected from the system. We give theoretical results detailing how one can transfer interventional information across time steps and define a dynamic causal GP model which can be used to quantify uncertainty and find optimal interventions in practice. We demonstrate how DCBO identifies optimal interventions faster than competing approaches in multiple settings and applications.

References in corpus (14)

• Counterfactual Multi-Agent Policy Gradients
• Directed Cyclic Graphical Representations of Feedback Models
• Learning Contextual Bandits in a Non-stationary Environment
• Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search
• On Deducing Conditional Independence from d-Separation in Causal Graphs with Feedback (Research Note)
• Time-Varying Gaussian Process Bandit Optimization
• Deconfounding Reinforcement Learning in Observational Settings
• From Ordinary Differential Equations to Structural Causal Models: the deterministic case
• Causal interpretation of stochastic differential equations
• Causal Discovery and Forecasting in Nonstationary Environments with State-Space Models
• From Deterministic ODEs to Dynamic Structural Causal Models
• Causal Inference using Gaussian Processes with Structured Latent Confounders
• Bayesian Optimization for Dynamic Problems
• Dynamic Batch Bayesian Optimization