Confidence intervals for multiple isotonic regression and other monotone models

Abstract

We consider the problem of constructing pointwise confidence intervals in the multiple isotonic regression model. Recently, Han and Zhang (2020) obtained a pointwise limit distribution theory for the so-called block max–min and min–max estimators (Fokianos, Leucht and Neumann (2020); Deng and Zhang (2020)) in this model, but inference remains a difficult problem due to the nuisance parameter in the limit distribution that involves multiple unknown partial derivatives of the true regression function.

In this paper, we show that this difficult nuisance parameter can be effectively eliminated by taking advantage of information beyond point estimates in the block max–min and min–max estimators. Formally, let $\hat{\mathit{u}}({\mathit{x}}_0)$ (resp. $\hat{\mathit{v}}({\mathit{x}}_0)$) be the maximizing lower-left (resp. minimizing upper-right) vertex in the block max–min (resp. min–max) estimator, and ${\hat{\mathit{f}}}_{\mathit{n}}$ be the average of the block max–min and min–max estimators. If all (first-order) partial derivatives of ${\mathit{f}}_0$ are nonvanishing at ${\mathit{x}}_0$, then the following pivotal limit distribution theory holds:

$$\sqrt{{\mathit{n}}_{\hat{\mathit{u}},\hat{\mathit{v}}}({\mathit{x}}_0)}({\hat{\mathit{f}}}_{\mathit{n}}({\mathit{x}}_0)-{\mathit{f}}_0({\mathit{x}}_0))⇝\mathit{\sigma }·{\mathbb{L}}_{{\mathbf{1}}_{\mathit{d}}}.$$

Here ${\mathit{n}}_{\hat{\mathit{u}},\hat{\mathit{v}}}({\mathit{x}}_0)$ is the number of design points in the block $[\hat{\mathit{u}}({\mathit{x}}_0),\hat{\mathit{v}}({\mathit{x}}_0)]$, σ is the standard deviation of the errors, and ${\mathbb{L}}_{{\mathbf{1}}_{\mathit{d}}}$ is a universal limit distribution free of nuisance parameters. This immediately yields confidence intervals for ${\mathit{f}}_0({\mathit{x}}_0)$ with asymptotically exact confidence level and oracle length. Notably, the construction of the confidence intervals, even new in the univariate setting, requires no more efforts than performing an isotonic regression once using the block max–min and min–max estimators, and can be easily adapted to other common monotone models including, for example, (i) monotone density estimation, (ii) interval censoring model with current status data, (iii) counting process model with panel count data, and (iv) generalized linear models. Extensive simulations are carried out to support our theory.