FEW-GROUP CROSS SECTIONS MODELING BY ARTIFICIAL NEURAL NETWORKS

This work deals with the modeling of homogenized few-group cross sections by Artificial Neural Networks (ANN). A comprehensive sensitivity study on data normalization, network architectures and training hyper-parameters specifically for Deep Shallow Feed Forward ANN is presented. The optimal models in terms reduction library size time are compared to multi-linear interpolation a Cartesian grid. use case provided OECD-NEA Burn-up Credit Criticality Benchmark [1]. Pytorch [2] machine learning framework used.