Conditionality principle under unconstrained randomness

Vladimir Vovk

Conditionality principle under unconstrained randomness

Research output: Contribution to journal › Article › peer-review

Abstract

A very simple example demonstrates that Fisher's application of the conditionality principle to regression ("fixed-x regression"), endorsed by David Sprott and many other followers, makes prediction impossible in the context of statistical learning theory. On the other hand, relaxing the requirement of conditionality makes it possible via, e.g., conformal prediction.

Original language	English
Number of pages	4
Journal	Statistical Science
Publication status	Accepted/In press - 8 Mar 2025

Keywords

assumption of randomness
conditionality principle
machine learning
prediction
regression

Access to Document

https://arxiv.org/abs/2402.04859

Embargoed Document

Accepted Manuscript
Accepted author manuscript, 109 KB

Cite this

@article{367a70d316ae4f7399391963561248c0,

title = "Conditionality principle under unconstrained randomness",

abstract = "A very simple example demonstrates that Fisher's application of the conditionality principle to regression ({"}fixed-x regression{"}), endorsed by David Sprott and many other followers, makes prediction impossible in the context of statistical learning theory. On the other hand, relaxing the requirement of conditionality makes it possible via, e.g., conformal prediction.",

keywords = "assumption of randomness, conditionality principle, machine learning, prediction, regression",

author = "Vladimir Vovk",

year = "2025",

month = mar,

day = "8",

language = "English",

journal = "Statistical Science",

issn = "0883-4237",

publisher = "Institute of Mathematical Statistics",

}

TY - JOUR

T1 - Conditionality principle under unconstrained randomness

AU - Vovk, Vladimir

PY - 2025/3/8

Y1 - 2025/3/8

N2 - A very simple example demonstrates that Fisher's application of the conditionality principle to regression ("fixed-x regression"), endorsed by David Sprott and many other followers, makes prediction impossible in the context of statistical learning theory. On the other hand, relaxing the requirement of conditionality makes it possible via, e.g., conformal prediction.

AB - A very simple example demonstrates that Fisher's application of the conditionality principle to regression ("fixed-x regression"), endorsed by David Sprott and many other followers, makes prediction impossible in the context of statistical learning theory. On the other hand, relaxing the requirement of conditionality makes it possible via, e.g., conformal prediction.

KW - assumption of randomness

KW - conditionality principle

KW - machine learning

KW - prediction

KW - regression

M3 - Article

SN - 0883-4237

JO - Statistical Science

JF - Statistical Science

ER -