Asymptotics for sliced average variance estimation

TL;DR

This paper analyzes the asymptotic properties of sliced average variance estimation (SAVE), revealing its sensitivity to the number of slices and proposing bias correction for improved consistency, especially with continuous responses.

Contribution

It provides a theoretical comparison between SAVE and sliced inverse regression (SIR), highlighting SAVE's limitations and proposing bias correction for better asymptotic behavior.

Findings

SAVE is not $\

and cannot achieve $\

when each slice has a fixed number of data points; bias correction improves its consistency; discretization allows $\

Abstract

In this paper, we systematically study the consistency of sliced average variance estimation (SAVE). The findings reveal that when the response is continuous, the asymptotic behavior of SAVE is rather different from that of sliced inverse regression (SIR). SIR can achieve $\sqrt{n}$ consistency even when each slice contains only two data points. However, SAVE cannot be $\sqrt{n}$ consistent and it even turns out to be not consistent when each slice contains a fixed number of data points that do not depend on n, where n is the sample size. These results theoretically confirm the notion that SAVE is more sensitive to the number of slices than SIR. Taking this into account, a bias correction is recommended in order to allow SAVE to be $\sqrt{n}$ consistent. In contrast, when the response is discrete and takes finite values, $\sqrt{n}$ consistency can be achieved. Therefore, an approximation through discretization, which is commonly used in practice, is studied. A simulation study is carried out for the purposes of illustration.