From conformal to probabilistic prediction

TL;DR

This paper introduces a novel probabilistic prediction method derived from conformal prediction principles, demonstrating promising results on the USPS dataset.

Contribution

It presents a new probabilistic prediction approach based on conformal prediction, extending existing methods with practical application.

Findings

Effective probabilistic predictions on USPS data

Encouraging empirical results

Potential for broader application

Abstract

This paper proposes a new method of probabilistic prediction, which is based on conformal prediction. The method is applied to the standard USPS data set and gives encouraging results.