An Improved Global Risk Bound in Concave Regression

TL;DR

This paper improves the theoretical risk bounds for concave function estimation using least squares, removing unnecessary logarithmic factors and extending results to model misspecification.

Contribution

It establishes a tighter risk bound of order n^{-4/5} for concave regression, eliminating the logarithmic factor from previous bounds.

Findings

Risk bound scales as n^{-4/5} without logarithmic factors.

Bound holds in expectation and with high probability.

Extension to model misspecification cases.

Abstract

A new risk bound is presented for the problem of convex/concave function estimation, using the least squares estimator. The best known risk bound, as had appeared in \citet{GSvex}, scaled like $\log(en) n^{-4/5}$ under the mean squared error loss, up to a constant factor. The authors in \cite{GSvex} had conjectured that the logarithmic term may be an artifact of their proof. We show that indeed the logarithmic term is unnecessary and prove a risk bound which scales like $n^{-4/5}$ up to constant factors. Our proof technique has one extra peeling step than in a usual chaining type argument. Our risk bound holds in expectation as well as with high probability and also extends to the case of model misspecification, where the true function may not be concave.