Loading...

Bayesian regularization of multilayer perceptron neural network for estimation of mass attenuation coefficient of gamma radiation in comparison with different supervised model-free methods

Moshkbar Bakhshayesh, K ; Sharif University of Technology | 2020

762 Viewed
  1. Type of Document: Article
  2. DOI: 10.1088/1748-0221/15/11/P11019
  3. Publisher: IOP Publishing Ltd , 2020
  4. Abstract:
  5. Multilayer perceptron (MLP) neural networks have been used extensively for estimation/regression of parameters. Moreover, recent studies have shown that learning algorithms of MLP which are based on Gaussian function are more accurate. In this paper, the mass attenuation coefficient (MAC) of gamma radiation for light-weight materials (e.g. O-8), mid-weight materials (e.g. Al-13), and heavy-weight materials (e.g. Pb-82) is modelled using Gaussian function based regularization of MLP (i.e. Bayesian regularization (BR)) and by a modular estimator. The results are compared with the Reference results. To show better performance of the utilized algorithm, the results of the different supervised methods including support vector machine (SVM) with different kernel functions, decision tree (DT), and radial basis network (RBN) are given. Average mean relative error (AMRE) and cumulative distribution function (CDF) of errors of MACs estimation are calculated. Comparison of the results indicates that MLP-BR gives more accurate results (e.g. AMREO−8 = 0.0014, CDFO−8(0.0069) = 0.99, AMREAl−13 = 0.0015, CDFAl−13(0.0048) = 0.99, AMREPb−82 = 0.0117, CDFPb−82(0.0523) = 0.99). c 2020 IOP Publishing Ltd and Sissa Medialab
  6. Keywords:
  7. Calibration ; Cluster finding ; Fitting methods ; Pattern recognition ; Decision trees ; Distribution functions ; Gamma rays ; Learning algorithms ; Multilayers ; Support vector machines ; Trees (mathematics) ; Bayesian regularization ; Cumulative distribution function ; Lightweight materials ; Mass attenuation coefficients ; Mean relative error ; Multilayer perceptron neural networks ; Radial basis networks ; Supervised methods ; Multilayer neural networks ; Interaction of radiation with matter ; Radiation calculations
  8. Source: Journal of Instrumentation ; Volume 15, Issue 11 , November , 2020
  9. URL: https://iopscience.iop.org/article/10.1088/1748-0221/15/11/P11019