Non parametric estimation of causal populations in a counterfactual scenario
This addresses a major issue in causality for researchers and practitioners, but appears incremental as it builds on existing methods like auto-encoders.
The paper tackles the problem of estimating causal treatment effects without confounding by reformulating it as a missing data model to estimate hidden causal populations, achieving this through a Causal Auto-Encoder with a prior and mask constraints.
In causality, estimating the effect of a treatment without confounding inference remains a major issue because requires to assess the outcome in both case with and without treatment. Not being able to observe simultaneously both of them, the estimation of potential outcome remains a challenging task. We propose an innovative approach where the problem is reformulated as a missing data model. The aim is to estimate the hidden distribution of \emph{causal populations}, defined as a function of treatment and outcome. A Causal Auto-Encoder (CAE), enhanced by a prior dependent on treatment and outcome information, assimilates the latent space to the probability distribution of the target populations. The features are reconstructed after being reduced to a latent space and constrained by a mask introduced in the intermediate layer of the network, containing treatment and outcome information.