Title :
Information-theoretic approach to unimodal density estimation
Author :
Brockett, Patrick L. ; Charnes, A. ; Paick, Kwang H.
Author_Institution :
Center for Cybern. Studies, Texas Univ., Austin, TX, USA
fDate :
5/1/1995 12:00:00 AM
Abstract :
Extends the maximum entropy information-theoretic density estimation method to provide a technique which guarantees that the resulting density is unimodal. The method inputs data in the form of moment or quantile constraints and consequently can handle both data-derived and non-data-derived information
Keywords :
maximum entropy methods; parameter estimation; probability; data-derived information; information-theoretic approach; maximum entropy; moment constraints; nondata-derived information; probability density distribution; quantile constraints; unimodal density estimation; Bayesian methods; Computer science; Cybernetics; Data engineering; Entropy; Estimation theory; Information analysis; Kernel; Probability density function; Statistical distributions;
Journal_Title :
Information Theory, IEEE Transactions on
