A note on the inverse estimator for the linear calibration problem

Abstract

In the simple linear calibration problem, two main competing estimators are the classical estimator and the inverse estimator. The Bayesian and compound-estimation approaches have been used to support the inverse estimator. The purpose of this note is to show that one can also derive the inverse estimator by the cross-validatory method. This derivation does not require any specific distributional assumptions.

Keywords:

• Linear calibration
• Inverse estimator
• Cross-Validation method