Two Principles for Data Based Probabilistic System Analysis

by Vicente Solana, Natl Research Council of Spain, Spain,
Niels C. Lind, Natl Research Council of Spain, Spain,

Abstract: A major problem in probabilistic system analysis is to find a self-consistent method of inference about probabilities based on a random sample of system state variables. Such a method is formalized as a two-phase process. Invariance requirements are reviewed and expressed as a strong invariance principle. A principle of data monotonicity is restated on the basis of propositional probability formalism. Distributions may be assigned by minimization of cross-entropy using fractile constraints. It is shown that this method satisfies the requirements of both principles.

Subject Headings: Probability | System analysis | Failure analysis | Entropy methods | Data analysis | Structural systems | Structural analysis | Structural failures

Services: Buy this book/Buy this article


Return to search