A Ubiquitous Method For Predicting Underground Petroleum Deposits Based On Satellite Data

This article investigates the use of machine learning methods to extract properties from satellite data for the purpose of detecting petroleum deposits. The authors build training and testing datasets and apply a number of machine learning methods to each of the training datasets. The best-performing models are then applied to the corresponding testing datasets and the results are used to build a matrix based on the performances of the best-performing models.