Timezone: »

 
Learning Counterfactually Invariant Predictors
Cecilia Casolo · Krikamol Muandet

@
in Algorithmic Fairness through the Lens of Causality and Privacy »

We propose a method to learn predictors that are invariant under counterfactual changes of certain covariates. This method is useful when the prediction target is causally influenced by covariates that should not affect the predictor output. For instance, this could prevent an object recognition model from being influenced by position, orientation, or scale of the object itself. We propose a model-agnostic regularization term based on conditional kernel mean embeddings to enforce \counterfactual invariance during training. We prove the soundness of our method, which can handle mixed categorical and continuous multivariate attributes. Empirical results on synthetic and real-world data demonstrate the efficacy of our method in a variety of settings.

[ Poster] 

Author Information

Cecilia Casolo (Technische Universität München)
Krikamol Muandet (Max Planck Institute for Intelligent Systems)

More from the Same Authors