Estimation of shear wave velocity from post-stack seismic data through committee machine with cuckoo search optimized intelligence models

Shear wave velocity (VS) afford petroleum engineers to favorable information for implementing of geomechanical, geophysical, and reservoir characterization studies. Hence, it is all-important to present a scheme for estimation of this parameter. The primary objective of this study is introducing a novel strategy for determining shear wave velocity from seismic data. For achieving the aforementioned objective, four-steps procedure are followed in this study: (1) Suitable seismic attributes (SAs) are selected as independent variables to estimate shear wave velocity by using step-wise regression method; (2) Input variables which selected in the first step are transformed into higher correlated data space through alternation conditional expectation (ACE); (3) Quantitative formulation between shear wave velocity and ACE transformed of input variables is made through three improved models namely optimized neural network (ONN), optimized support vector regression (OSVR), and optimized fuzzy inference system (OFIS). Optimization implementation of intelligence models are achieved through cuckoo search method (CS); (4) Committee machine (CM) using cuckoo search method is employed for integrating optimized models so reaps those benefits. The efficiency of proposed models is assessed founded on statistical parameters. The obtained results corroborate the superb performance of committee machine in preference to its elements. This paper infers the proposed strategy is suitable for modeling of shear wave velocity as a function of seismic data.