Understanding shale gas and tight oil productivity with machine learning

I show how the accuracy of production decline curve forecasting–a widely used but ill-posed inverse problem–can be improved by using a Bayesian formulation and machine learning.

Please contact author for more info: jbm@mit.edu