Neutron-Gamma Classification by Evolutionary Fuzzy Rules and Support Vector Machines

Accurate and fast methods for neutron-gamma discrimination play an essential role in the development of digital scintillation detectors. Digital detectors allow the use of state-of-the-art data analysis, mining, and classification methods in place of traditional approaches based on analog technology such as the pulse rise-time and charge-comparison methods. This work compares the ability of evolutionary fuzzy rules and support vector machines to perform accurate neutron-gamma classification. The accuracy and performance of both investigated methods are evaluated on two real-world data sets.