A reliable version of choquistic regression based on evidence theory

Choquistic regression is an elegant generalisation of logistic regression, which preserves its monotonicity whilst alleviating its linearity. However, much as logistic regression, it lacks self-awareness, that is, an ability to represent the ignorance ( aka epistemic uncertainty) involved in its predictions, which is crucial in safety-critical classification problems. Recently, an extension of logistic regression was introduced to remedy this issue for this latter classifier. This extension is formalised within evidence theory and relies in particular on a sound method for statistical inference and prediction developed in this framework. In this paper, a similar extension is derived for choquistic regression. The usefulness of the obtained approach is confirmed empirically in classification problems where cautiousness in decision-making is allowed.