Chapter Ten

Statistical Inference with Minimum Relative Entropy

A Robust Numerical Algorithm Employing Sinc Quadrature

Vasilios G. Koures1    IISAM L3C, Cheyenne, WY, USA1 Corresponding author: email address: koures@iisam.com

Abstract

Given partial information (i.e., constraints) about a probability distribution, the distribution that maximizes the entropy with respect to the constraints is the one that is least prejudiced about the missing information. As new information arrives, the prior distribution must change. To maintain maximum uncertainty given new information, we must minimize the relative entropy (the Kullback–Leibler distance) between the prior (p0) and posterior (p) distributions. Robust constrained optimization algorithms to ...

Get Concepts of Mathematical Physics in Chemistry: A Tribute to Frank E. Harris - Part A now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.