Neural Modeling of Shear Wave Sonic Logs

F010 Neural Modeling of Shear Wave Sonic Logs Abstract 1 In this paper we show a neural network based nonlinear regression to develop a relationship between shear wave log data and a set of other well logs to predict pseudo shear wave logs in a well that does not contain shear wave information. We use a special network architecture in combination with parallel learning automatic network growing and sophisticated stopping criteria that allows an efficient and robust estimation of the optimal network size for this particular problem. Initially we use gamma-ray two resistivity and the p-wave sonic logs as model