The Models of Determination of Spin Values from Experimental Properties of Nuclei on the Base of Other Experimental Properties

Abstract

In December 2012 the group of physicists from Atomic Mass Data Center, located at Centre de Spectrométrie Nucléaire et de Spectrométrie de Masse (CSNSM), Orsay, France, shared the databases of atomic nuclei (cf.[1]) containing a lot of experimental data characterizing the nucleus, for example, nuclear ground-state masses and radii, magnetic moments, half-lives, spins and parities of excited and ground-state, their decay modes and the relative intensities of these decays, the deformations and many others. Today there are about 2830 nuclei that have been observed. It is estimated that the total number of nuclei which may be obtained by experimenters is between 6000 and 7000. It is evident that many properties of atomic nuclei are so far unknown and we usually define them using a variety of theoretical methods, based on those properties of nuclei which are already known. One of the most important properties is the spin of the nucleus in the ground state, specifying also the possible excited states. The main aim of the presented paper is to use the latest experimental data to check the existing theoretical methods of the prediction of unknown spins of nuclei and proposals for new estimates based on the methods of data mining and artificial intelligence [4,6,8,9,11] We compare the different models of the prediction of spin value with the help of the R programistic language, well fitted to statistical and data mining modelling. The properties of atomic nuclei have been collected in the MySQL database. In order to integrate the R language with MySQL we use the package RMySQL which contains the interface to communicate with the MySQL.