Kremena Arsova–Borisova, Veselin Hristov
ABSTRACT. The differences between shallow neural networks and deep neural networks are considered. Data from operational exploration of an open pit mine are used to train different types of deep neural networks to predict a useful indicator. The results of deep network training are compared both with each other and with previous results obtained from the training of shallow neural networks. The errors in training and testing of the models according to the number of layers and nodes in the networks are analysed.
Key words: deep neural networks, deep learning
Arsova-Borisova, K., V. Hristov
DEEP NEURAL NETWORKS APPLICATIONS IN THE STUDY OF A GEOLOGICAL INDICATOR DOI: 10.5281/zenodo.5594817
DOI: https://doi.org/10.58903/b15171913
References
1. Han J., Kamber M., Pei J., 2011. Data Mining: Concepts and Techniques, Third Edition – Elsevier Inc., USA.
2. Topalov St., Hristov V., Prognostication of Geological Parameters Value Using Neural Networks, Proceedings of International Society for Mine Surveying XIII International Congress, Budapest, Hungary, 24-28 September 2007, No 048. Elected from Institute of Mine Surveyors of South Africa Journal, XXXIV (2), 2010, 18 – 25.
3. Topalov St., Hristov V., Investigation of 3D copper grade changeability by Neural Networks in metasomatic ore deposit, Proceedings of the 9th WSEAS International Conferences on: NEURAL NETWORKS (NN’08), May 2-4, 2008, Sofia, Bulgaria, 58 – 62.
4. Hristov V., Topalov St., Influence of geological information density on the training quality of neural networks as forecast models, Proceedings of the 6th Jordanian International Mining Conference, 01 – 03 November, 2011, Amman. Jordan
5. Hristov V., Topalov St., Varianti na prognozirane s nevronni mrezhi pri podzemen dobiv na olovno – cinkova ruda, Proceedings of the Treta nacionalna nauchno – tehnicheska konferencija s mezhdunarodno uchastie „Tehnologii i praktiki pri podzemnija dobiv i minnoto stroitelstvo”, 8 – 11 October, 2012, Devin, Bulgaria, 26 – 31. (in Bulgarian)
6. Hristov V., 2015. Nevronni mrezhi pri analiz na danni ot minnata promishlenost, publishing house MGU “Sv. Ivan Rilski”, Sofia, 108 p (in Bulgarian)
7. Hristov V., Topalov St., Tochnost na prognozirane na kachestven pokazatel s nevronni mrezhi obuslovena na krajnija kontur na raboten horizont, Proceedings of the XІII-th National Conference with International Participation of the Open and Underwater Mining of Minerals, 1 – 5 September 2015, Varna, Bulgaria, 221 – 226. (In Bulgarian)