A Novel Application of Ensemble Methods with Data Resampling Techniques for Drill Bit Selection in the Oil and Gas Industry

Selection of the most suitable drill bit type is an important task for drillers when planning new oil and gas wells. With advancement intelligent predictive models, automated selection possible using earlier drilled offset wells’ data. However, real-field well data samples naturally involve unequal distribution points that results in formation a complex imbalance multi-class classification problem during selection. In this analysis, Ensemble methods, namely Adaboost Random Forest, have been combined with re-sampling techniques to provide approach handling process. Additionally, four popular machine learning namely, K-nearest neighbors, naïve Bayes, multilayer perceptron, support vector machine, are also evaluated understand performance degrading effects imbalanced drilling obtained from Norwegian The comparison shows random forest bootstrap class weighting technique has given impressive testing accuracy ranges 92% 99%, G-mean (0.84–0.97) critical normal experimental scenarios. This study provides automate process over any field, which will minimize human error, time, cost.