Title :
Embedding maximum entropy models in algebraic varieties by Gröbner bases methods
Author :
Dukkipati, Ambedkar
Author_Institution :
Dept. of Comput. Sci. & Autom., Indian Inst. of Sci., Bangalore, India
fDate :
June 28 2009-July 3 2009
Abstract :
The main aim of this paper is to present some notions on how results from commutative algebra and algebraic geometry could be used in representation and computation of maximum entropy (ME) models in the cases, where an integer valued sufficient statistic exists. We give an implicit description of ME-models by embedding them in algebraic varieties for which we use Grobner bases methods. We prove that in the case of ME, both the model and the data can be represented by algebraic varieties.
Keywords :
algebra; geometry; maximum entropy methods; Grobner bases methods; algebraic geometry; algebraic varieties; commutative algebra; maximum entropy; Algebra; Computational geometry; Entropy; Equations; Lagrangian functions; Parametric statistics; Polynomials; Probability distribution; Random variables; Solid modeling;
Conference_Titel :
Information Theory, 2009. ISIT 2009. IEEE International Symposium on
Conference_Location :
Seoul
Print_ISBN :
978-1-4244-4312-3
Electronic_ISBN :
978-1-4244-4313-0
DOI :
10.1109/ISIT.2009.5205528