Full text loading...
-
Bayesian Lithology/Fluid Prediction Base on Efficient Kernel Density Estimation
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 79th EAGE Conference and Exhibition 2017, Jun 2017, Volume 2017, p.1 - 5
Abstract
We present a LF estimation method based on Bayesian setting. The key problems in the inversion method based on Bayesian framework is the approximation of the conditional probability density. Fast kernel density estimate method is used to compute the likelihood function and this method maintains more statistical characteristic between the parameters than Gaussian assumption, thus this method can improved the precision of the inversion method. The probabilities of LF classes are estimated based on GMM and KDE, and compared their advantages, disadvantages, and applicability. The mentioned method can quantitatively describe the uncertainty in the process of LF prediction, and it’s of great significance in seismic interpretation.