Corpus ID: 211029267

Copy, paste, infer: A robust analysis of twin networks for counterfactual inference

  title={Copy, paste, infer: A robust analysis of twin networks for counterfactual inference},
  author={L. Graham and Ciar{\'a}n M. Lee},
Twin networks are a simple method for estimating counterfactuals, originally proposed to have several advantages over standard counterfactual inference. However, no study yet exists exploring in what contexts twin networks would be more advantageous than standard counterfactual methods in practice. We conduct an empirical and theoretical analysis of twin networks to show that in certain cases of Structural Causal Models, twin networks are faster and less memory intensive by orders of magnitude… Expand

Figures from this paper

Estimating the probabilities of causation via deep monotonic twin networks
There has been much recent work using machine learning to answer causal queries. Most focus on interventional queries, such as the conditional average treatment effect. However, as noted by Pearl,Expand


Learning Representations for Counterfactual Inference
A new algorithmic framework for counterfactual inference is proposed which brings together ideas from domain adaptation and representation learning and significantly outperforms the previous state-of-the-art approaches. Expand
Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search
The Counterfactually-Guided Policy Search (CF-GPS) algorithm is proposed, which leverages structural causal models for counterfactual evaluation of arbitrary policies on individual off-policy episodes and can improve on vanilla model-based RL algorithms by making use of available logged data to de-bias model predictions. Expand
Causal Responsibility and Counterfactuals
A novel model of intuitive judgments of responsibility is proposed that is a function of both pivotality (whether an agent made a difference to the outcome) and criticality (how important the agent is perceived to be for the outcome, before any actions are taken). Expand
What Counterfactuals Can Be Tested
This paper provides a complete characterization of "testable counterfactuals," namely,Counterfactual statements whose probabilities can be inferred from physical experiments, and provides complete procedures for discerning whether a givencounterfactual is testable and, if so, expressing its probability in terms of experimental data. Expand
Counterfactual Fairness
This paper develops a framework for modeling fairness using tools from causal inference and demonstrates the framework on a real-world problem of fair prediction of success in law school. Expand
Counterfactual diagnosis
This work reformulates diagnosis as a counterfactual inference task and derive new counterfactUAL diagnostic algorithms that are closer to the diagnostic reasoning of clinicians and significantly improves the accuracy and safety of the resulting diagnoses. Expand
Estimating individual treatment effect: generalization bounds and algorithms
A novel, simple and intuitive generalization-error bound is given showing that the expected ITE estimation error of a representation is bounded by a sum of the standard generalized-error of that representation and the distance between the treated and control distributions induced by the representation. Expand
Reliable Decision Support using Counterfactual Models
This work proposes using a different learning objective that predicts counterfactuals instead of predicting outcomes under an existing action policy as in supervised learning, and introduces the Counterfactual Gaussian Process (CGP) to support decision-making in temporal settings. Expand
Probabilistic Evaluation of Counterfactual Queries
A formalism that uses probabilistic causal networks to evaluate one's belief that the counterfactual consequent, C, would have been true if the antecedent, A, were true is presented. Expand
Probabilities Of Causation: Three Counterfactual Interpretations And Their Identification
  • J. Pearl
  • Mathematics, Computer Science
  • Synthese
  • 2004
It is shown thatnecessity and sufficiency are two independent aspects of causation, and that both should be invoked in the construction of causal explanations for specific scenarios. Expand