TY - JOUR
T1 - Error control and loss functions for the deep learning inversion of borehole resistivity measurements
AU - Shahriari, Mostafa
AU - Pardo, David
AU - Rivera, Jon A.
AU - Torres-Verdín, Carlos
AU - Picon, Artzai
AU - Del Ser, Javier
AU - OSSANDON VELIZ, SEBASTIAN EDUARDO
AU - Calo, Victor M.
N1 - Publisher Copyright:
© 2020 John Wiley & Sons, Ltd.
Copyright:
Copyright 2021 Elsevier B.V., All rights reserved.
PY - 2021/3/30
Y1 - 2021/3/30
N2 - Deep learning (DL) is a numerical method that approximates functions. Recently, its use has become attractive for the simulation and inversion of multiple problems in computational mechanics, including the inversion of borehole logging measurements for oil and gas applications. In this context, DL methods exhibit two key attractive features: (a) once trained, they enable to solve an inverse problem in a fraction of a second, which is convenient for borehole geosteering operations as well as in other real-time inversion applications. (b) DL methods exhibit a superior capability for approximating highly complex functions across different areas of knowledge. Nevertheless, as it occurs with most numerical methods, DL also relies on expert design decisions that are problem specific to achieve reliable and robust results. Herein, we investigate two key aspects of deep neural networks (DNNs) when applied to the inversion of borehole resistivity measurements: error control and adequate selection of the loss function. As we illustrate via theoretical considerations and extensive numerical experiments, these interrelated aspects are critical to recover accurate inversion results.
AB - Deep learning (DL) is a numerical method that approximates functions. Recently, its use has become attractive for the simulation and inversion of multiple problems in computational mechanics, including the inversion of borehole logging measurements for oil and gas applications. In this context, DL methods exhibit two key attractive features: (a) once trained, they enable to solve an inverse problem in a fraction of a second, which is convenient for borehole geosteering operations as well as in other real-time inversion applications. (b) DL methods exhibit a superior capability for approximating highly complex functions across different areas of knowledge. Nevertheless, as it occurs with most numerical methods, DL also relies on expert design decisions that are problem specific to achieve reliable and robust results. Herein, we investigate two key aspects of deep neural networks (DNNs) when applied to the inversion of borehole resistivity measurements: error control and adequate selection of the loss function. As we illustrate via theoretical considerations and extensive numerical experiments, these interrelated aspects are critical to recover accurate inversion results.
KW - deep learning
KW - deep neural networks
KW - error estimation
KW - geophysical applications
KW - real-time inversion
UR - http://www.scopus.com/inward/record.url?scp=85099013066&partnerID=8YFLogxK
U2 - 10.1002/nme.6593
DO - 10.1002/nme.6593
M3 - Article
AN - SCOPUS:85099013066
VL - 122
SP - 1629
EP - 1657
JO - International Journal for Numerical Methods in Engineering
JF - International Journal for Numerical Methods in Engineering
SN - 0029-5981
IS - 6
ER -