A COMPARATIVE STUDY ON APPLICABILITY AND EFFICIENCY OF MACHINE LEARNING ALGORITHMS FOR MODELING GAMMA-RAY SHIELDING BEHAVIORS

다국어 초록 (Multilingual Abstract)

The mass attenuation coefficient is the primary physical parameter to model narrow beam gamma-rayattenuation. A new machine learning based approach is proposed to model gamma-ray shieldingbehavior of composites alternative to theoretical calculations. Two fuzzy logic algorithms and a neuralnetwork algorithm were trained and tested with different mixture ratios of vanadium slag/epoxy resin/antimony in the 0.05 MeVe2 MeV energy range. Two of the algorithms showed excellent agreementwith testing data after optimizing adjustable parameters, with root mean squared error (RMSE) valuesdown to 0.0001. Those results are remarkable because mass attenuation coefficients are often presentedwith four significant figures. Different training data sizes were tried to determine the least number ofdata points required to train sufficient models. Data set size more than 1000 is seen to be required tomodel in above 0.05 MeV energy. Below this energy, more data points with finer energy resolution mightbe required. Neuro-fuzzy models were three times faster to train than neural network models, whileneural network models depicted low RMSE. Fuzzy logic algorithms are overlooked in complex functionapproximation, yet grid partitioned fuzzy algorithms showed excellent calculation efficiency and goodconvergence in predicting mass attenuation coefficient.

참고문헌 (Reference)

1 M. Berger, "XCOM: Photon Cross Sections Database"

2 Huseyin Ozan Tekin ; Viswanath P. Singh ; Tugba Manici ; Elif Ebru Altunsoy, "Validation of MCNPX with Experimental Results of Mass Attenuation Coefficients for Cement, Gypsum and Mixture" 대한방사선방어학회 42 (42): 154-157, 2017

3 O. Klein, "Uber die Streuung von Strahlung durch freielektronen nach der neuen relativistischen quantendynamik von Dirac" 52 : 853-868, 1928

4 O. Gencel, "The application of artificial neural networks technique to estimate mass attenuation coefficient of shielding barrier" 4 (4): 743-751, 2009

5 M. Sugeno, "Structure identification of fuzzy model" 28 (28): 15-33, 1988

6 A. G. Bakirtzis, "Short term load forecasting using fuzzy neural networks" 10 (10): 1518-1524, 1995

7 J H Hubbell, "Review and history of photon cross section calculations" IOP Publishing 51 (51): R245-R262, 2006

8 J. K. Shultis, "Radiation Shielding and Radiological Protection in Handbook of Nuclear Engineering" Springer 2010

9 F. Rosenblatt, "Principles of Neurodynamics. Perceptrons and the Theory of Brain Mechanisms" Cornell Aeronautical Lab Inc 1961

10 I. Akkurt, "Prediction of photon attenuation coefficients of heavy concrete by fuzzy logic" 347 (347): 1589-1597, 2010

11 H. Bethe, "On the stopping of fast particles and on the creation of positive electrons" 146 (146): 83-112, 1934

12 J.J. Mor e, "Numerical Analysis" Springer 105-116, 1978

13 R.H. Pratt, "New Relativistic S-Matrix Results for Scattering beyond the Usual Anomalous Factors/beyond Impulse Approximation" Lawrence Livermore National Lab 1993

14 C. M. Bishop, "Neural Networks for Pattern Recognition" Oxford university press 1995

15 S. Alp, "Modelling of multi-objective transshipment problem with fuzzy goal programming" 6 : 9-20, 2018

16 N. Kucuk, "Modeling of gamma ray energy-absorption buildup factors for thermoluminescent dosimetric materials using multilayer perceptron neural network : a comparative study" 86 : 10-22, 2013

17 The Math Works, Inc, "MATLAB"

18 P. Goyal, "Large Mini Batch Sgd: Training Image Net in 1 Hour"

19 F. H. Attix, "Introduction to Radiological Physics and Radiation Dosimetry" John Wiley & Sons 2008

20 A.S. Lapedes, "How Neural Nets Work"

21 L.A. Zadeh, "Fuzzy sets" Elsevier BV 8 (8): 338-353, 1965

22 G. Zhang, "Forecasting with artificial neural networks : the state of the art" 14 (14): 35-62, 1998

23 V. P. Singh, "Determination of mass attenuation coefficient for some polymers using Monte Carlo simulation" 119 : 284-288, 2015

24 J. Hamilton, "Coulomb corrections in non-relativistic scattering" 60 : 443-477, 1973

25 L.J. Herrera, "Clustering-Based TSK neuro-fuzzy model for function approximation with interpretable sub-models" Springer 1007-, 2005

26 A. Davydenko, "Business Forecasting: Practical Problems and Solutions" Wiley 2016

27 M. E. Medhat, "Application of neural network for predicting photon attenuation through materials" 3-4 (3-4): 171-181, 2019

28 E. E. Zadeh, "Application of artificial neural network in precise prediction of cement elements percentages based on the neutron activation analysis" 131 (131): 167-, 2016

29 A. Yadollahi, "Application of artificial neural network for predicting the optimal mixture of radiation shielding concrete" 89 : 69-77, 2016

30 H. B. Kavanoz, "A novel comprehensive utilization of vanadium slag/epoxy resin/antimony trioxide ternary composite as gamma ray shielding material by MCNP 6.2 and BXCOM" 165 : 108446-, 2019

31 J. C. F. Pujol, "A neural network approach to fatigue life prediction" 33 (33): 313-322, 2011