Discovery of Mineralization Predication Classification Rules by Using Gene Expression Programming Based on PCA

Abstract

Classification is one of the fundamental tasks in geology field. In this paper, we propose an evolutionary approach for discovering classification rules of mineralization predication from distinct combinations of geochemistry elements by using gene expression programming (GEP). The innovative part of the paper presents integrated/hybrid model-combine GEP evolution modeling with Principal Component Analysis (PCA), which reduce multidimensional data sets. Mineral deposit with tin and copper in Gejiu is chosen as the research area. MAPGIS and MORPAS are used to extract the value of ore-controlled factors by mapping geologic maps into grid cell. Case study illustrates the proposed GEP approach Based on PCA is more efficient and accurate in a large searching space, compared with Decision Tree (C4.5) and Bayesian Networks.