Artificial neural network data analysis for classification of soils based on their radionuclide content
Abstract
The artificial neural network (ANN) data analysis method was used to recognize and classify soils of an unknown geographic origin. A total of 103 soil samples were differentiated into classes according to the regions in Serbia and Montenegro from which they were collected. Their radionuclide (Ra-226, U-238, U-235, K-40, Cs-134, Cs-137, Th-232, and Be-7) activities detected by gamma-ray spectrometry were then used as inputs to ANN. Five different training algorithms with different numbers of samples in training sets were tested and compared in order to find the one with the minimum root mean square error (RMSE). The best predictive power for the classification of soils from the fifteen regions was achieved using a network with seven hidden layer nodes and 2500 training epochs using the online back-propagation randomized training algorithm. With the optimized ANN, most soil samples not included in the ANN training data set were correctly classified at an average rate of 92%.
Source:
Russian Journal of Physical Chemistry A, 2007, 81, 9, 1477-1481Publisher:
- Maik Nauka/Interperiodica/Springer, New York
Funding / projects:
DOI: 10.1134/S0036024407090257
ISSN: 0036-0244
WoS: 000253706400025
Scopus: 2-s2.0-34548494799
Institution/Community
Tehnološko-metalurški fakultetTY - JOUR AU - Dragović, Snežana D. AU - Onjia, Antonije PY - 2007 UR - http://TechnoRep.tmf.bg.ac.rs/handle/123456789/1182 AB - The artificial neural network (ANN) data analysis method was used to recognize and classify soils of an unknown geographic origin. A total of 103 soil samples were differentiated into classes according to the regions in Serbia and Montenegro from which they were collected. Their radionuclide (Ra-226, U-238, U-235, K-40, Cs-134, Cs-137, Th-232, and Be-7) activities detected by gamma-ray spectrometry were then used as inputs to ANN. Five different training algorithms with different numbers of samples in training sets were tested and compared in order to find the one with the minimum root mean square error (RMSE). The best predictive power for the classification of soils from the fifteen regions was achieved using a network with seven hidden layer nodes and 2500 training epochs using the online back-propagation randomized training algorithm. With the optimized ANN, most soil samples not included in the ANN training data set were correctly classified at an average rate of 92%. PB - Maik Nauka/Interperiodica/Springer, New York T2 - Russian Journal of Physical Chemistry A T1 - Artificial neural network data analysis for classification of soils based on their radionuclide content EP - 1481 IS - 9 SP - 1477 VL - 81 DO - 10.1134/S0036024407090257 ER -
@article{ author = "Dragović, Snežana D. and Onjia, Antonije", year = "2007", abstract = "The artificial neural network (ANN) data analysis method was used to recognize and classify soils of an unknown geographic origin. A total of 103 soil samples were differentiated into classes according to the regions in Serbia and Montenegro from which they were collected. Their radionuclide (Ra-226, U-238, U-235, K-40, Cs-134, Cs-137, Th-232, and Be-7) activities detected by gamma-ray spectrometry were then used as inputs to ANN. Five different training algorithms with different numbers of samples in training sets were tested and compared in order to find the one with the minimum root mean square error (RMSE). The best predictive power for the classification of soils from the fifteen regions was achieved using a network with seven hidden layer nodes and 2500 training epochs using the online back-propagation randomized training algorithm. With the optimized ANN, most soil samples not included in the ANN training data set were correctly classified at an average rate of 92%.", publisher = "Maik Nauka/Interperiodica/Springer, New York", journal = "Russian Journal of Physical Chemistry A", title = "Artificial neural network data analysis for classification of soils based on their radionuclide content", pages = "1481-1477", number = "9", volume = "81", doi = "10.1134/S0036024407090257" }
Dragović, S. D.,& Onjia, A.. (2007). Artificial neural network data analysis for classification of soils based on their radionuclide content. in Russian Journal of Physical Chemistry A Maik Nauka/Interperiodica/Springer, New York., 81(9), 1477-1481. https://doi.org/10.1134/S0036024407090257
Dragović SD, Onjia A. Artificial neural network data analysis for classification of soils based on their radionuclide content. in Russian Journal of Physical Chemistry A. 2007;81(9):1477-1481. doi:10.1134/S0036024407090257 .
Dragović, Snežana D., Onjia, Antonije, "Artificial neural network data analysis for classification of soils based on their radionuclide content" in Russian Journal of Physical Chemistry A, 81, no. 9 (2007):1477-1481, https://doi.org/10.1134/S0036024407090257 . .