Posterior contraction in group sparse logit models for categorical responses
share
This paper studies posterior contraction in multi-category logit models with priors incorporating group sparse structures. We provide a unified platform for contraction rates in high-dimensional logit models that subsume binary logistic regression under individual sparsity. No size restriction is directly imposed on the true signal in our study. In addition to establishing first-ever contraction properties for multi-category logit models under group sparsity, this work also refines the recent findings on the Bayesian theory of binary logistic regression.
READ FULL TEXT
