Comparer des méthodes

	Estimation bayésienne de la densité par noyau	Krigeage bayésien (géostatistique basée sur un modèle)
Domaine	Analyse spatiale	Analyse spatiale
Famille	Regression model	Regression model
Année d'origine	1995	1993–1998
Auteur d'origine	Hjort & Glad (1995); extended by various authors in Bayesian nonparametrics	Diggle, Tawn & Moyeed; Handcock & Stein
Type	Nonparametric density estimation	Bayesian spatial interpolation
Source fondatrice	Hjort, N. L., & Glad, I. K. (1995). Nonparametric density estimation with a parametric start. The Annals of Statistics, 23(3), 882–904.	Diggle, P. J., Tawn, J. A., & Moyeed, R. A. (1998). Model-based geostatistics. Journal of the Royal Statistical Society: Series C (Applied Statistics), 47(3), 299–350.
Alias	Bayesian KDE, BKDE, Bayesian nonparametric density estimation, Bayesian adaptive KDE	Bayesian geostatistics, model-based geostatistics, Bayesian spatial interpolation, stochastic kriging
Apparentées	5	5
Résumé	Bayesian Kernel Density Estimation (BKDE) is a nonparametric method for estimating the probability density function of a spatial or attribute variable by combining a kernel smoother with a Bayesian prior over the bandwidth parameter. The posterior distribution of the bandwidth propagates uncertainty into the final density estimate rather than treating the bandwidth as a fixed tuning constant.	Bayesian Kriging embeds classical geostatistical interpolation inside a full probabilistic framework. Instead of treating variogram parameters as fixed point estimates, it places prior distributions on them and updates these priors with observed spatial data to obtain a posterior distribution. Predictions at unsampled locations are then marginalised over this uncertainty, yielding honest predictive intervals that account for both spatial dependence and parameter uncertainty.