CAVIAR: Categorical-Variable Embeddings for Accurate and Robust Inference

arXiv:2404.04979v1 Announce Type: cross
Abstract: Social science research often hinges on the relationship between categorical variables and outcomes. We introduce CAVIAR, a novel method for embedding categorical variables that assume values in a high-dimensional ambient space but are sampled from an underlying manifold. Our theoretical and numerical analyses outline challenges posed by such categorical variables in causal inference. Specifically, dynamically varying and sparse levels can lead to violations of the Donsker conditions and a failure of the estimation functionals to …

abstract ambient arxiv categorical challenges cs.lg econ.em embedding embeddings inference manifold novel numerical relationship research robust science social social science space type values variables