LSTM-Based Deep Learning Method for Automated Detection of Geophysical Signatures in Mining

Abstract The mining of stratified ore deposits requires detailed knowledge the location orebody boundaries. In Banded Iron Formation (BIF) hosted iron located in Pilbara region Western Australia natural gamma logs are useful tool to identify stratigraphic However, manually interpreting these features is subjective and time consuming due large volume data. this study, we propose a novel approach automatically detect signatures. We implemented LSTM based algorithm for automated detection achieved relatively high accuracy using sequences with without added noise. Further, no feature extraction or selection performed work. Hence, can be used different signatures even So, system introduced as an aid geoscientists.