What-If Reasoning with Counterfactual Gaussian Processes

By Peter F. Schulam, Suchi Saria. 2017

8 citations recorded by the source corpus.

Answering "What if?" questions is important in many domains. For example, would a patient's disease progression slow down if I were to give them a dose of drug A? Ideally, we answer our question using an experiment, but this is not always possible (e.g., it may be unethical). As an alternative, we can use non-experimental data to learn models that make counterfactual predictions of what we would observe had we run an experiment. In this paper, we propose the counterfactual GP, a counterfactual model of continuous-time trajectories (time series) under sequences of actions taken in continuous-time. We develop our model within the potential outcomes framework of Neyman and Rubin. The counterfactual GP is trained using a joint maximum likelihood objective that adjusts for dependencies between observed actions and outcomes in the training data. We report two sets of experimental results using the counterfactual GP. The first shows that it can be used to learn the natural progression (i.e. untreated progression) of biomarker trajectories from observational data. In the second, we show how the CGP can be used for medical decision support by learning counterfactual models of renal health under different types of dialysis.

Subfields: Management Science and Operations Research, Probabilistic & Bayesian Methods, Causal Inference & Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Management Science and Operations Research

1. Ala Al-Fuqaha
2. Junaid Qadir
3. Suchi Saria
4. Hossein Soleimani
5. Adarsh Subbaswamy

In Probabilistic & Bayesian Methods

1. Ala Al-Fuqaha
2. Junaid Qadir
3. Suchi Saria
4. Hossein Soleimani
5. Adarsh Subbaswamy

In Causal Inference & Discovery

1. Adarsh Subbaswamy
2. Suchi Saria
3. Hossein Soleimani
4. Bryan Lim

Citations in the corpus

• Neural Ordinary Differential Equations (2018, neutral): “, 2018b) and can naturally model continuous-time effects and interventions (Soleimani et al., 2017b; Schulam and Saria, 2017).”
• Secure and Robust Machine Learning for Healthcare: A Survey (2020, background): “For instance, the Gaussian processes based counterfactual causal model has been presented in [189] and in”
• Learning Treatment-Response Models from Multivariate Longitudinal Data (2017, background): “…worth noting that in the potential outcomes framework, estimation of the causal effect is precluded by the need to justify a set of generally untestable assumptions (Gill and Robins, 2001); e.g., see Lok (2008); Schulam and Saria (2017) for assumptions needed for continuoustime potential outcomes.”