Estimation of thermal neutron cross sections from K, U, Th concentrations for rock samples using neural network algorithms

In the paper present the results of the use of the neural network algorithms to find a function Σa=f(K, U, Th,...). The easily measurable parameters (K, U, Th concentrations, lithology) were used to estimate the thermal neutron absorption cross-section Σa, which is difficult to measure in the borehole conditions. This paper is suggesting a possible solution to the problem. This method may have an important application in the well-logging data treatment. (author). 6 refs, 9 tabs