Prediction of Coal-Bed Methane Production Based on PCA and SVM

Accurate prediction of oil production is difficult because there are so many factors affecting production. In this paper, new screening rules are used for principal component analysis and factor screening, the support vector machine is used to complete subsequent learning and prediction work, and Gaussian regression is selected to determine the dividing line of prediction results. The results show that the model is more accurate in predicting Wells with production below 2000 cubic meters, and more accurate in predicting Wells with production above 2000 cubic meters.