Sonic Waves Travel-time Prediction: When Machine Learning Meets Geophysics

Sonic wave travel-time prediction is an important task in oil and gas exploration as it provides important information on the content and lithography of the rocks. Travel-time data, however, are not always accessible due to practical considerations. Currently, machine learning methods have been used to infer these values. In this paper, we look at the application of machine learning in predicting sonic wave travel-time, specifically in terms of challenges, benchmarks, and datasets. In addition, we present some preliminary results of sonic wave travel-time prediction using existing machine learning regression methods, namely curve fitting artificial neural network and multiple linear regression. Finally, this paper is aimed to act as a "bridge" between machine learning practitioners and domain-specific oil and gas engineers.