Research On Mineral Resources Evaluation Based On Data Mining Technology

Mineral resources is one of the foundations of the country’s sustained economicdevelopment, and it needs to find new mineral resources along with the increasing of the crisismines. The development of earth sciences provides a wealth of geological data for theevaluation of mineral resources. But geoscience data has large, high-dimensional, incomplete,noise, fuzzy features which bring some difficulties to the traditional methods of resourceevaluation. In this case, data mining technology provides an effective solution for the problem.The traditional geochemical prospecting method is to draw the contour maps which baseson the sampling of each sampling point within the region and each chemical element content ofvalue, then analyze and compare the contour maps. By this means, heavy workload, lowefficiency, the evaluation results are affected by the artificial subjective factors.The difference between the traditional geochemical prospecting and our ways in this paperlies: We introduced the cluster analysis algorithm in the data mining technology for dataclustering, and determined the resource occurrence target; Then used correlation analysis todetermine the target-occurrence of resource types in the data mining technology, it provided anew theoretical method for the evaluation of mineral resources.Firstly，the geochemical sampling points are selected as the data objects and the16kindsof measuring elements are used as the attributes of the data objects, The paper researched andanalyzed the distribution characteristics of geochemical elements in survey area by clusteringthe sampling points with K-means arithmetic and Cluster arithmetic about data miningtechnology. In the researched area, the cluster analysis results showed the two geochemicalelement combination of characteristics consistent with the sampling-area(The properties of thedata object changing characteristics is identical). The clustering results and the regionalgeological data are superimposed and analysed, and determined a target area of occurrence ofmineral resources. A target is located in the K-feldspar granite and granite plaque area;Another target area is located in andesitic basalt, andesite, basaltic andesite area.After that, we researched the target area of mineral resources using correlation analysisthat analyzed the relationship of the geochemical elements(a symbiotic characteristics of thegeochemical elements), the first target is the correlation of W and Bi, Sb and As, Cu and Co,Coand Ni, Ni and Cu, Zn and of Cd, Cd and Ag; The second target is the correlation of As、Sb、Hg、Mo，Pb、Cd，W、Ni. Form geochemical theory we can deduct that the first target is tofind the favorable area of the silver, copper, lead, zinc polymetallic ore; The second target is to find a favorable areas for gold, silver, lead, zinc. Through looking up the relevant articles，theevaluation and the inferred results of the paper are similar to now-known researchachievements of the region.It shows that data mining technology can be effectively applied tothe evaluation of mineral resources, can greatly improve work efficiency, and be able to get agood evaluation results.