Title :
The latent maximum entropy principle
Author :
Wang, Shaojun ; Rosenfeld, Ronald ; Zhao, Yunxin ; Schuurmans, Dale
Author_Institution :
Sch. of Comput. Sci., Carnegie Mellon Univ., Pittsburgh, PA, USA
Abstract :
We present an extension of Jaynes´ maximum entropy principle to handle latent variables. We use an EM algorithm that incorporates nested iterative scaling to approximately calculate maximum entropy solutions for this principle, and give a proof of its convergence.
Keywords :
convergence of numerical methods; information theory; iterative methods; maximum entropy methods; EM algorithm; Jaynes maximum entropy principle; convergence; latent variables; nested iterative scaling; Concurrent computing; Entropy; Geographic Information Systems; Iterative algorithms; Power system modeling; Probability distribution; Statistics;
Conference_Titel :
Information Theory, 2002. Proceedings. 2002 IEEE International Symposium on
Print_ISBN :
0-7803-7501-7
DOI :
10.1109/ISIT.2002.1023403
