Mineral Potential Prediction Using Hybrid Intelligent Approach

Predicting minerals potential helps miners in making wise decisions. In this paper, we present a computerized prototype system called GoldXplorer that enables to predict the existence and distribution of gold (aurum (Au)) at a given location with its geological and geographical factors. With its two intelligent engines based on back-propagation neural network and self-organizing map techniques, GoldXplorer helps users (geologists, investors, gold miners or individuals) to determine the gold potentials for further investigation and mining activity. For the case study, we use data sets collected from an area called Kuala Lipis in Malaysia. This set of data was supplied and verified by Malaysia Department of Minerals and Geosciences. The data consists of the location (RSOE and RSON) and all types of minerals found in the area. Two thousand sample points were used in this research and it results in eighty percent (or more) of accuracy rate. With the results produced by GoldXplorer, the potential strategic locations for gold mining may be determined. In addition, with the predicted information, future planning can be carried out so that the earth can be properly kept from any illegal mining.