CBM Gas Content Prediction Model Based on the Ensemble Tree Algorithm with Bayesian Hyper-Parameter Optimization Method: A Case Study of Zhengzhuang Block, Southern Qinshui Basin, North China

Gas content is an important parameter for evaluating coalbed methane reservoirs, so it prerequisite resource evaluation and favorable area optimization to predict the gas accurately. To improve accuracy of CBM prediction, Bayesian hyper-parameter method (BO) introduced into random forest algorithm (RF) gradient boosting decision tree (GBDT) establish prediction models using well-logging data in Zhengzhuang block, south Qinshui Basin, China. As a result, GBDT model based on BO (BO-GBDT model) RF (BO-RF were proposed. The results show that mean-square-error (MSE) BO-RF BO-GBDT can be reduced by 8.83% 37.94% average less than modes, indicating optimized improved. effect better model, especially low wells, R-squared (RSQ) 0.82 0.66. order different was > RF. Compared with other models, curve predicted has best fitness measured content. rule distribution all four consistent distribution.