Akademik Veri Yönetim Sistemi
Radiation Physics and Chemistry, cilt.208, 2023 (SCI-Expanded)
In the studies of nuclear physics, one of the important parameters for nuclear reactions is the reaction cross-section. It can be obtained from experimental data or by different theoretical models. In this study, we implement machine learning (ML) algorithms to the regression analysis of the nuclear cross-section of neutron-induced nuclear reactions of Germanium isotopes. The data for the training of the machine was borrowed from the TENDL-2019 library for the total cross-section data of possible nuclear reactions after the bombardment of the different target materials by neutrons. Three ML models, Artificial Neural Networks (ANN), K-Nearest Neighbors (KNN) and, Support Vector Machines (SVM), were developed to fit nuclear data from the TENDL-2019 database in order to predict neutron induce reaction cross sections. The performance of each algorithm is determined and compared by evaluating the mean square error (MSE) and the correlation coefficient (R2). According to the results obtained, we demonstrate that cross-section information can be obtained safely with ML techniques and the regression curve generated by our models is in good agreement with the evaluated nuclear data library. From our study, ANN and KNN are found to be better compared to SVM algorithm. ML models can enhance classical physics-guided models and play a role in nuclear data analyses. They can be used as an alternative to the estimation of cross-sections for neutron energies of an unknown energy value.