A convex formulation for high-dimensional sparse sliced inverse regression

TL;DR

This paper introduces a convex optimization approach for sparse sliced inverse regression in high-dimensional settings, enabling direct subspace estimation and variable selection to improve interpretability and accuracy.

Contribution

It proposes a novel convex formulation that estimates the subspace and performs variable selection simultaneously in high-dimensional sliced inverse regression.

Findings

Successfully identifies relevant covariates in simulations

Provides theoretical bounds on subspace estimation accuracy

Demonstrates improved interpretability in high-dimensional data

Abstract

Sliced inverse regression is a popular tool for sufficient dimension reduction, which replaces covariates with a minimal set of their linear combinations without loss of information on the conditional distribution of the response given the covariates. The estimated linear combinations include all covariates, making results difficult to interpret and perhaps unnecessarily variable, particularly when the number of covariates is large. In this paper, we propose a convex formulation for fitting sparse sliced inverse regression in high dimensions. Our proposal estimates the subspace of the linear combinations of the covariates directly and performs variable selection simultaneously. We solve the resulting convex optimization problem via the linearized alternating direction methods of multiplier algorithm, and establish an upper bound on the subspace distance between the estimated and the true subspaces. Through numerical studies, we show that our proposal is able to identify the correct covariates in the high-dimensional setting.