Open AccessApplication of clusterization algorithms for building materials classification on radioactivity in R
Author(s) -
P. E. Sokolov,
S. A. Sentenberg
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/687/2/022003
Subject(s) - raw material , raw data , computer science , segmentation , radionuclide , basis (linear algebra) , cluster (spacecraft) , algorithm , data mining , artificial intelligence , mathematics , chemistry , nuclear physics , physics , organic chemistry , programming language , geometry
The clusterization method of raw construction materials on their radioactivity with clusterization algorithms within statistical treatment environment and R analysis applied is handled in the paper. The authors make an emphasis on application of modern methods of raw construction materials radioactivity data treatment. Raw construction materials are evaluated on the basis of objective evidence, reflecting indices of specific activities and of inner natural radionuclides specific activity. As a result of net learning on the experimental data the authors acquire a cluster map provided with segmentation on the natural radionuclides effective specific activity. Consequently, both the criteria are established and the hitting of a raw material into the corresponding cluster is estimated. The application of a “decision tree” algorithm makes it possible to state rules according to which raw materials refer to these or those clusters. As opposed to traditional analysis methods, the applied technique of radioactivity estimation is based on quantitative characteristics. On the ground of the acquired results conclusions are made on a possibility and appropriateness of application of the said technique of analysis and classification of raw materials radioactivity data.