Bayesian Triangle Smoothing
Ivan Sendin1; Thiago Queiroz2; Marcos Batista2
1 UFU; 2 UFG
doi:10.20906/CPS/USM-2016-0024
Resumo
Triangle bound smoothing is used to reduce the uncertainty from a set of Euclidean distances: an iterative process uses triangle inequality to remove the slack on those distances, providing a new set of distances. In this work we introduce a Bayesian version of triangle bound smoothing that also produce statistical information on those distances. The method was applied to a set of distances that simulate a protein on an NMR experiment and the resulting set of distances obtained a mean squared error value lower than the initial
Palavras-chave: Bayesian methods