Inductive Venn-Abers and related regressors

TL;DR

This paper extends Venn-Abers predictors from binary classification and bounded regression to unbounded regression by incorporating conformal prediction, improving predictive efficiency in larger training datasets.

Contribution

It introduces a generalization of Venn-Abers predictors to unbounded regression, combining conformal prediction to enhance applicability and efficiency.

Findings

Venn-Abers regressors can be effectively extended to unbounded regression.

Point regressors derived from Venn-Abers improve predictive efficiency for larger datasets.

Simulation and empirical studies support the effectiveness of the proposed method.

Abstract

Venn-Abers predictors are probabilistic predictors that enjoy appealing properties of validity, but their major limitation is that they are applicable only to the case of binary classification, with a recent extension to bounded regression. We generalize them to the case of unbounded regression, which requires adding an element of conformal prediction. In our simulation and empirical studies we investigate the predictive efficiency of point regressors derived from Venn-Abers regressors and argue that they somewhat improve the predictive efficiency of standard regressors for larger training sets.