Selective sampling after solving a convex problem

TL;DR

This paper develops a model-agnostic, change-of-measure approach for selective inference after solving convex statistical learning problems, providing explicit Jacobian formulas and sampling methods for complex penalties.

Contribution

It introduces a general change-of-measure formula and geometric analysis for selective inference post convex optimization, including non-polyhedral penalties like group LASSO.

Findings

Derived explicit Jacobian formulas for group LASSO.

Established a change-of-measure framework for various penalties.

Proposed a projected Langevin sampler for log-concave distributions.

Abstract

We consider the problem of selective inference after solving a (randomized) convex statistical learning program in the form of a penalized or constrained loss function. Our first main result is a change-of-measure formula that describes many conditional sampling problems of interest in selective inference. Our approach is model-agnostic in the sense that users may provide their own statistical model for inference, we simply provide the modification of each distribution in the model after the selection. Our second main result describes the geometric structure in the Jacobian appearing in the change of measure, drawing connections to curvature measures appearing in Weyl-Steiner volume-of-tubes formulae. This Jacobian is necessary for problems in which the convex penalty is not polyhedral, with the prototypical example being group LASSO or the nuclear norm. We derive explicit formulae for the Jacobian of the group LASSO. To illustrate the generality of our method, we consider many examples throughout, varying both the penalty or constraint in the statistical learning problem as well as the loss function, also considering selective inference after solving multiple statistical learning programs. Penalties considered include LASSO, forward stepwise, stagewise algorithms, marginal screening and generalized LASSO. Loss functions considered include squared-error, logistic, and log-det for covariance matrix estimation. Having described the appropriate distribution we wish to sample from through our first two results, we outline a framework for sampling using a projected Langevin sampler in the (commonly occuring) case that the distribution is log-concave.