Distance anomaly factors for gold potential mapping in the Jinchanggouliang area, Inner Mongolia, China

• Published in: Earth science informatics Vol. 14; no. 2; pp. 1083 - 1099
• Main Authors: Chen, Yongliang, Sun, Guosheng, Zhao, Qingying
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.06.2021; Springer Nature B.V
• Subjects: Earth and Environmental Science; Earth Sciences; Earth System Sciences; Euclidean geometry; Gold; Information Systems Applications (incl.Internet); Kernel functions; Mapping; Neural networks; Ontology; Optimization; Polynomials; Radial basis function; Research Article; Simulation and Modeling; Space Exploration and Astronautics; Space Sciences (including Extraterrestrial Physics; Support vector machines; Unit cell; China; Inner Mongolia China; Gold potential mapping; Bat algorithm; Distance anomaly factor; One-class support vector machine; Receiver operating characteristic curve; Area under the curve
• ISSN: 1865-0473; 1865-0481
• DOI: 10.1007/s12145-021-00614-5

Distance anomaly factors (DAFs) were defined for each cell of the unit cell population in a study area to represent mineral potential of the cell. A DAF of a cell is a formula for the total distance from the cell to all other cells in the study area. The distance between two cells can be expressed as the Manhattan distance, Canberra distance, Euclidean distance, and kernel Euclidean distance. The kernels in the kernel Euclidean distance can be radial basis function (RBF) kernel, chi-squared kernel, sigmoid kernel, Laplacian kernel, and polynomial kernel. Accordingly, eight DAFs were defined to map gold potential of the Jinchanggouliang area, Inner Mongolia, China. The receiver operating characteristic (ROC) curve analysis was used to evaluate the effectiveness of the eight DAFs. The results show that these DAFs are comparable to one-class support vector machine (OCSVM) in gold potential mapping. The optimal threshold for distinguishing gold potential cells from all the cells was determined by maximizing the Youden index. The gold potential targets predicted by the eight DAFs occupy 7.4% – 16.5% of the study area, while containing 78% - 91% of the discovered gold deposits. The gold potential targets predicted by the default parameter OCSVM and by the bat-optimized OCSVM occupy 9.4% and 16.4% of the study area, respectively, while containing 78% and 87% of the discovered gold deposits. Therefore, the eight DAFs are feasible approaches for gold potential mapping. Their effectiveness needs to be further tested in mineral potential mapping in other areas.