Neural network regression analysis and post-stack inversion- A comparison
Somanath Misra, Satinder Chopra, Yingsheng Zhang · 2010
Probabilistic Neural Network(PNN) is used to perform a multi-attribute regression analysis to compute P-wave velocity and density parameters over a small seismic data volume acquired in Alberta, Canada. The estimated density and P-wave velocity parameters are subsequently used to compute the P-impedance volume for the seismic data volume. We also compute the P-impedance volume from the post-stack data using the standard model-based inversion technique. We compare the results obtained with the neural network regression analysis and standard inversion and show that both the approaches corroborate well with each other.