Conformal predictive distributions with kernels

TL;DR

This paper reviews the evolution of predictive distributions in statistics, highlights recent advances integrating them with machine learning and kernel methods, and introduces new developments in this area.

Contribution

It introduces a novel approach combining predictive distributions with kernel methods, expanding their application in machine learning.

Findings

Predictive distributions are effectively integrated with kernel methods.

Recent literature has advanced the application of predictive distributions in machine learning.

The paper provides a historical perspective on the development of predictive distributions.

Abstract

This paper reviews the checkered history of predictive distributions in statistics and discusses two developments, one from recent literature and the other new. The first development is bringing predictive distributions into machine learning, whose early development was so deeply influenced by two remarkable groups at the Institute of Automation and Remote Control. The second development is combining predictive distributions with kernel methods, which were originated by one of those groups, including Emmanuel Braverman.