Machine learning in radioactive nuclides identification

Chemistry and nuclear physics represent one of possible options of large area for practical data mining applications. Radio-active nuclides are related with many sectors, such as medicine or industry. Successful detection and identification of radio-nuclides allows the realization of specific safety precaution and then increasing of security level in nuclear power plants, or medicine institutes. Data obtained by radio spectroscopy method reach high level of relevancy. Thus these data are suitable for using data mining methods, because the presented problem corresponds with classification task. Two radio-nuclides identification methods were presented in this paper; the second method is designed specially for this purpose.