Use Machine Learning to Classify Materials Based on Gamma Scattering Spectra
Huynh Thanh Nhan, Nguyen Duy Thong, Le Hoang Minh, Tran Thien Thanh, Chau Van Tao
IEEJ Trans 2025
Abstract:
In this study, machine learning is used to determine materials and thickness of materials based on gamma scattering spectra. Materials used in this study are: Al, Si, Fe, Mn, Mg, Co, Cu, Zn, and Ti, which have thicknesses varying from 1 mm to 50 mm. In order to estimate thickness as well as material simultaneously, 1-scattering spectrum and 2-scattering spectrum are used. The Random Forest algorithm was used in training and evaluating the machine learning model. Results of this study provided a coefficient of determination R2 = 0.990 and mean squared error MSE = 1.250.
More detail >>
