Foundations of Probability Theory for AI - The Application of Algorithmic Probability to Problems in Artificial Intelligence

TL;DR

This paper introduces Algorithmic Complexity Theory as a foundation for probability and demonstrates its application to AI problems, offering near-optimal search strategies for broad problem classes.

Contribution

It presents the application of Algorithmic Probability to AI, providing a theoretical basis for near-optimal search methods in complex problem domains.

Findings

Algorithmic Complexity defines probability with characteristic properties.

Application of Algorithmic Probability yields near-optimal search procedures.

Potential for broad classes of AI problems to benefit from this approach.

Abstract

This paper covers two topics: first an introduction to Algorithmic Complexity Theory: how it defines probability, some of its characteristic properties and past successful applications. Second, we apply it to problems in A.I. - where it promises to give near optimum search procedures for two very broad classes of problems.