Speaker
Description
Regression models and artificial neural networks (ANNs) are among the most commonly employed techniques for predicting the dynamics of natural and industrial phenomena [1]. In eastern Sudan, radioactivity levels in areas affected by gold mining and other industrial activities were forecast using multiple regression and artificial neural networks. We developed multiregression and ANN models using Python scripts on a Linux-based computer.Using the findings of background radiation measurements conducted in an area of naturally occurring radioactive materials (NORM) in Eastern Sudan, we tested and validated the model.The outcomes were contrasted with matching activity concentrations of soil samples determined by HP-Ge gamma spectrometry [2]. The mean and percentage variations between the measured and expected values of 40 K, 232Th, and 238U were analyzed. Artificial neural networks (ANNs) and multilinear regression were employed to estimate the activities of 232Th and 40 K, along with other radiological parameters. These results were compared to those obtained by gamma spectroscopy (GS) measurements. The expected values for radioactivity and radiological parameters were within the uncertainties of the measured values for the models. With an accuracy of up to 96%, the ANN surpassed the linear regression model in predicting radioactivity concentration and other targeted radiological risk indicators using ambient conditions and location coordinates as input. However, ANNs require significantly more computational power than regression models.
Keywords: NORM; Natural Radioactivity, gamma spectrometry; python programming, artificial intelligence, machine learning approaches.
References
1. Mathew, P. Amudha, and S. Sivakumari, ‘Deep learning techniques: an overview’, Adv. Intell. Syst. Comput., vol. 1141, no. January, pp. 599–608, 2021, doi: 10.1007/978-981-15-3383-9_54.
2. International Atomic Energy Agency (IAEA) and International Atomic Energy Agency (IAEA), 2011. Radiation protection and NORM residue management in the production of rare earths from thorium containing minerals. International Atomic Energy Agency.
