Research Article Open Access

Formalizing Probability Concepts in a Type Theory

Farida Kachapova1
  • 1 Auckland University of Technology, New Zealand

Abstract

In this paper we formalize some fundamental concepts of probability theory such as the axiomatic definition of probability space, random variables and their characteristics, in the Calculus of Inductive Constructions, which is a variant of type theory and the foundation for the proof assistant COQ. In a type theory every term and proposition should have a type, so in our formalizations we assign an appropriate type to each object in order to create a framework where further development of formalized probability theory will be possible. Our formalizations are based on mathematical results developed in the COQ standard library; we use mainly the parts with logic and formalized real analysis. In the future we aim to create COQ coding for our formalizations of probability concepts and theorems. In this paper the definitions and some proofs are presented as flag-style derivations while other proofs are more informal.

Journal of Mathematics and Statistics
Volume 14 No. 1, 2018, 209-218

DOI: https://doi.org/10.3844/jmssp.2018.209.218

Submitted On: 22 September 2018 Published On: 12 November 2018

How to Cite: Kachapova, F. (2018). Formalizing Probability Concepts in a Type Theory. Journal of Mathematics and Statistics, 14(1), 209-218. https://doi.org/10.3844/jmssp.2018.209.218

  • 3,819 Views
  • 2,687 Downloads
  • 0 Citations

Download

Keywords

  • Type Theory
  • Kolmogorov's Axiomatics
  • Probability Theory
  • Calculus of Inductive Constructions
  • Flag-Style Derivation