Research on construction method of data-driven minerals prediction model

This is a new era of computing everywhere, software definition and data driven development.In the mineral prediction, compared with previous statistical methods, the advantage of the machine learning and deep learning algorithm is that it can be better to show the complex nonlinear relationship between the mineralized point and the spatial factors.This paper combines geology, geophysical exploration, geochemical exploration and remote sensing data, and uses three algorithms which are decision tree, support vector machine and convolution neural network to carry out mineral prediction work of comprehensive information.According to the sample data of Beishan area in Gansu province, it is found that the modeling method of decision tree and support vector machine is more suitable than that of convolution neural network.