Formalization of Measure Theory and Lebesgue Integration for Probabilistic Analysis in HOL

  • Authors:
  • Tarek Mhamdi;Osman Hasan;Sofiène Tahar

  • Affiliations:
  • Concordia University;Concordia University;Concordia University

  • Venue:
  • ACM Transactions on Embedded Computing Systems (TECS) - Special Issue on Modeling and Verification of Discrete Event Systems
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Dynamic systems that exhibit probabilistic behavior represent a large class of man-made systems such as communication networks, air traffic control, and other mission-critical systems. Evaluation of quantitative issues like performance and dependability of these systems is of paramount importance. In this paper, we propose a generalized methodology to formally reason about probabilistic systems within a theorem prover. We present a formalization of measure theory in the HOL theorem prover and use it to formalize basic concepts from the theory of probability. We also use the Lebesgue integration to formalize statistical properties of random variables. To illustrate the practical effectiveness of our methodology, we formally prove classical results from the theories of probability and information and use them in a data compression application in HOL.