K-Means Clustering and General Regression Neural Network Methods for Copper Mineralization probability in Chahar-Farsakh, Iran

Due to the efficiency of data mining science for analyzing and reviewing extensive data. especially geochemical data, essential methods and techniques such as the hierarchical method, K-Means method density-based methods, Cohennon method, and so forth, have been developed and utilized by numerous researchers for clustering. One of the most notable and widely used algorithms in the field of clustering is the K-Means algorithm. This algorithm divides the data into K clusters by emphasizing the distance criterion. This study focuses on applying this method according to lithogeochemical data taken from the 1:100,000 scale map of Chahar-Farsakh in South Khorasan province for the elements of copper, cobalt and nickel to the sampling coordinates. The optimal value of K was classified according to the desirability of the selection and the data, and thus the relationships between these elements in the range were determined. This was analyzed by changing the value of K from 3 to 15 criteria mentioned in each class to reveal the optimal K. According to the observations, the existence of a quadratic relationship with negative concavity between copper and cobalt elements, as well as a special exponential relationship between copper and nickel and a positive linear relationship between nickel and cobalt, were reported. Finally, considering the coordinates of the samples and the concentration of cobalt and nickel. the quantity of copper was predicted using a General Regression Neural Network (GRNN). The accuracy of this method was estimated to be 0.99 on training data and 0.76 on test data Therefore, using the proposed method (K-means Clustering and GRNN) in this paper, it is possible to examine the extent of changes in other elements in the analysis. Also, it is possible to make deeper and broader explorations via determining the relationship between the elements.