Multi-element geochemical anomaly recognition applying geologically-constrained convolutional deep learning algorithm with Butterworth filtering

Deep learning networks have extensively been used to detect multi-element geochemical anomalies linked to different mineral deposits in recent decade. Efficient detection of multi-element geochemical anomalies is a significant issue that should be done through capable processing methods. The two-dimensional (2D) convolutional neural networks (CNN) are a powerful subset of machine learning algorithms which can present magnificent conclusions. Because those have the ability to extract high-level features of complex inputs. But, there are several shortcomings in applying 2D images as input of the CNNs for multi-element geochemical anomaly mapping. This challenge can be solved designing one-dimensional (1D) CNN to convolve geochemical data table yielding unbiased maps produced. Also, great feature extraction ability of the CNNs can increase defining geological constraints and training their frameworks with frequency domain data which is contained superlative information. In this regard, a novel geologically-constrained convolutional deep learning (GCDL) algorithm has been developed to classify multi-element geochemical data table. Accordingly, constructed GCDL with frequency domain training data of the Cr mineralization in the Robat Sefid district, NE Iran, could demonstrate impressive results. Based on success-rate curves, produced map of frequency domain geochemical data has plotted 86.22% Cr occurrences via 28% corresponding area while produced map of spatial domain geochemical data has plotted 79.91% Cr occurrences via same corresponding area.