Machine Learning Approach for Gamma-ray Spectra Identification for Radioactivity Analysis

We have proposed a machine learning model for efficient gamma-ray spectrometry for environmental recovery from the Fukushima Daiichi Power Plant Accident. In the present study, we focus on a radioactive nuclide identification by the machine learning in a screening measurement. A simple deep neural network having two hidden layer is proposed, and the identification accuracy is achieved more than 95% for single gamma-ray spectra.