Optimization of multi-element geochemical anomaly recognition via swarm-intelligence support vector machine in the Takht-e Soleyman area, northwestern Iran

Hamid Sabbaghi ^* seyed hassan Tabatabaei Hassan Tabatabaei

• Isfahan University of Technology, Isfahan, Isfahan, Iran

Mineral exploration is becoming increasingly troublous because the level at which undiscovered mineral deposits can be found is progressively becoming deeper under barren cover. Therefore, detecting metal resources under barren cover is a significant key for industrial competitions. However, the application of optimized machine learning algorithms is critical for detecting the presence of undiscovered deposits under barren cover. One of the most significant issues in mineral exploration is detection of multi-element geochemical anomalies that indicate the presence of undiscovered mineral deposits barren cover. Recently, many machine learning procedures have been developed and employed to model and map multi-element geochemical anomalies whereby their important hyperparameters have been regulated generally through trial-and-error. However, employing swarm-intelligence optimization techniques reduces training time and assists to obtain more precise results. In this research, a known swarm-intelligence procedure called grasshopper optimization algorithm was executed to optimize known hyperparameters of the support vector machine (SVM) for identifying multi-element geochemical anomalies in the Takhte Soleyman district, NW Iran. The grasshopper-optimized support vector machine (GSVM) proved to be a vigorous approach for detecting multi-element geochemical anomaly and it can also be regarded for other geoscience applications. An optimized SVM algorithm using polynomial kernel function (PF) and radial basis kernel function (RBF) could result in multi-element geochemical anomaly models with accuracies above 95% in the shortest possible time without using trial-and-error.