1986IEEE Transactions on ReliabilityRequires access

Maximum Entropy and Reliability Distributions

S. Teitler, A. K. Rajagopal, K. L. Ngai

Open publisher page 23 citations

Abstract

An effort is made to show the relevance and usefulness of the principle of maximum entropy to reliability considerations. The constraints entering into the maximum entropy principle are identified as a class of sufficient statistics which determine the unknown parameters in the probability densities that occur in the most commonly used reliability models. In this way, the maximum entropy principle is shown to be completely compatible with prevailing practice in failure analysis. It is also pointed out that the differential entropy is equal to unity minus the expectation of the natural logarithm of the hazard rate. Maximization of the differential entropy is therefore equivalent to minimization of the expectation of the logarithm of the hazard rate. Behavior of the differential entropy under transformation of variable is used as an indicator of change or lack of change of conditions of failure.

About this research paper

What this paper is about

An effort is made to show the relevance and usefulness of the principle of maximum entropy to reliability considerations. The constraints entering into the maximum entropy principle are identified as a class of sufficient statistics which determine the unknown parameters in the probability densities that occur in the most commonly used reliability models. In this way, the maximum entropy principle is shown to be completely compatible with prevailing practice in failure analysis. It is also pointed out that the differential entropy is equal to unity minus the expectation of the natural logarithm of the hazard rate. Maximization of the differential entropy is therefore equivalent to minimization of the expectation of the logarithm of the hazard rate. Behavior of the differential entropy under transformation of variable is used as an indicator of change or lack of change of conditions of failure.

Why it matters

OpenAlex reports 23 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

An effort is made to show the relevance and usefulness of the principle of maximum entropy to reliability considerations. The constraints entering into the maximum entropy principle are identified as a class of sufficient statistics which determine the unknown parameters in the probability densities that occur in the most commonly used reliability models. In this way, the maximum entropy principle is shown to be completely compatible with prevailing practice in failure analysis. It is also pointed out that the differential entropy is equal to unity minus the expectation of the natural logarithm of the hazard rate. Maximization of the differential entropy is therefore equivalent to minimization of the expectation of the logarithm of the hazard rate. Behavior of the differential entropy under transformation of variable is used as an indicator of change or lack of change of conditions of failure.

Key concepts: Maximum entropy probability distribution, Mathematics, Principle of maximum entropy, Differential entropy, Logarithm, Maximum entropy thermodynamics, Maximum entropy spectral estimation, Entropy (arrow of time)

Related papers

Back to paper searchBrowse research topicsOriginal source
Maximum Entropy and Reliability Distributions — Research Paper | ScholarLens