Robust machine-learning workflow for subsurface geomechanical characterization and comparison against popular empirical correlations
作者：
Highlights：
• Recovery of fossil/ geothermal energy requires subsurface mechanical characterization.
• Robust machine learning workflow is proposed for synthesizing sonic travel time logs.
• New data preprocessing and model stability assessment ensures the robustness.
• Multivariate adaptive regression spline performs better than neural network.
• Multivariate adaptive regression spline performs better than empirical models.
摘要
•Recovery of fossil/ geothermal energy requires subsurface mechanical characterization.•Robust machine learning workflow is proposed for synthesizing sonic travel time logs.•New data preprocessing and model stability assessment ensures the robustness.•Multivariate adaptive regression spline performs better than neural network.•Multivariate adaptive regression spline performs better than empirical models.
论文关键词：Machine Learning,Geomechanical,Sonic,Oil and Gas,Neural Network
论文评审过程：Received 4 September 2020, Revised 9 March 2021, Accepted 22 March 2021, Available online 26 March 2021, Version of Record 5 April 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.114942
