Seismic Facies Classification Using Bayesian Networks
Yuan Gu, Peimin Zhu, Rong Hui, Hai Yang, Fanping Zeng · 2010
Summary This article illustrates a novel approach (based on Bayesian Network. BN ) for seismic facies classification. The approach, by integrated all available seismic attributes, could not only give the qualitative analysis of classification, but the quantitative analysis result of prediction probability of seismic facie. In addition, if we have more valuable experience from previous similar research, BN provide a useful tool for adding these experiences into BN Learning processing to improve the result of classification. For synthetic data with 30% random Gauss noise, the output of the method shows considerable agreement with the geologic model; similarly, for the real data application, the predicted facies match those results of previous research.