Quantitative Identification of Microfacies Based on ICA,PCA and SVM
Xuefeng Zhang · Well Logging Technology · 2011
Identifying sedimentary facies based on log data is one of the important and desiderated problems in oil field exploration and development.Microfacies characteristics are reflected from log data.Some features described the microfacies changes based on the combination of conventional logging data,the geological interpretation results and core data are extracted by the principal component analysis(PCA) and independent component analysis(ICA).Quantitative identification model of microfacies based on support vector machine can automaticly determine the sedimentary microfacies types of the non-cores information drills.Practical application shows that the proposed method is better than traditional SVM which without feature extraction in small sample cases,and the recognition rate after ICA feature extraction is better than that of PCA.