Isotonic regression in multi-dimensional spaces and graphs

DOI10.1214/20-AOS1947zbMATH Open1462.62208arXiv1812.08944OpenAlexW3110679391MaRDI QIDQ1996789FDOQ1996789

Publication date: 26 February 2021

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: In this paper we study minimax and adaptation rates in general isotonic regression. For uniform deterministic and random designs in

[0, 1]^{d}

with

d g e 2

and

N (0, 1)

noise, the minimax rate for the

e l l_{2}

risk is known to be bounded from below by

n^{- 1 / d}

when the unknown mean function

f

is nondecreasing and its range is bounded by a constant, while the least squares estimator (LSE) is known to nearly achieve the minimax rate up to a factor

(l o g n)^{g} a m m a

where

n

is sample size,

g a m m a = 4

in the lattice design and

g a m m a = m a x 9 / 2, (d^{2} + d + 1) / 2

in the random design. Moreover, the LSE is known to achieve the adaptation rate

(K / n)^{- 2 / d} {1 v e e l o g (n / K)}^{2 g a m m a}

when

f

is piecewise constant on

K

hyperrectangles in a partition of

[0, 1]^{d}

. Due to the minimax theorem, the LSE is identical on every design point to both the max-min and min-max estimators over all upper and lower sets containing the design point. This motivates our consideration of estimators which lie in-between the max-min and min-max estimators over possibly smaller classes of upper and lower sets, including a subclass of block estimators. Under a