Abstract

A novel approach based on the concept of hybrid neural network (HNN) has been used for classifying sediment types using downhole log data acquired during Integrated Ocean Drilling Program (IODP) Expedition 323 in the Bering Sea Slope region. Prior to applying supervised classification, the multi-variate based statistical analysis was carried out to know the cluster which supports the present analysis. The stochastic Markov Chain Monte Carlo (MCMC)/Hybrid Monte Carlo (HMC) learning paradigm has been used to constrain the lithology boundaries using density, density porosity, gamma ray, sonic P-wave velocity and electrical resistivity at the hole U1344A. The hybrid results based on HMC algorithm (BNN.HMC) resolves finer structures at certain depths in addition to main lithology such as silty clay, diatom clayey silt and sandy silt. The present analysis provides detail pictures of sediment information from a depth ranging between 615 to 655 meter below seafloor (mbsf) of no core recovery zone. The present analysis demonstrates that the hybrid approach renders robust means for the classification of complex lithology successions at the hole U1344A, which could be very useful for other studies and understanding the oceanic sediment inhomogeneity and structural discontinuities.