ABSTRACT
This paper uses the SVM machine learning method for stratigraphic division of logging curve, and takes advantage of the same region existing lamination well data change characteristics to serve as the breakthrough point. Combined with logging curve data of each well and representative characteristic of change about formation properties of logging curve data changes, the paper builds a reasonable mathematical model, realizes the automatic lamination. The practice showed that according to the actual problem with logical choice of 2-class of classification model and kernel functions, the SVM method is effective regarding the division of logging curve.
