Entropy maximization (Q1043868): Difference between revisions

In the paper it is shown that: (i) every probability density is the unique maximizer of the relative entropy in an appropriate class, and (ii) in the class of all probability density functions \(f\) that satisfy \(\int f\,h_id\mu=\lambda_i\) for \(i= 1,2,\dots, k,\) the maximizer of the entropy is an \(f_0\) that is proportional to \(\exp(\sum c_ih_i)\) for some choice of the \(c_i\). An extension of this to a continuum of constraints and 10 examples are presented. The references contain 5 bibliographical hints: \textit{R. Durrett}, Probability. Theory and examples. CA: Wadsworth (1991; Zbl 0709.60002); \textit{I. Karatzas} and \textit{S.E. Shreve}, Brownian motion and stochastic calculus. NY: etc.: Springer (1988; Zbl 0638.60065); \textit{H.P. McKean}, Stochastic integrals. NY-London: Acad. Press (1969; Zbl 0191.46603); \textit{W. Rudin}, Real and complex analysis. 3\,rd ed., NY: McGraw-Hill (1987; Zbl 0925.00005); and \textit{C.E. Shannon}, Bell Syst. Tech. J. 27, 379--423, 623--656 (1948; Zbl 1154.94303).