Imputation in well log data: A benchmark for machine learning methods

In this work, a review of the well log data imputation literature was presented, together with the proposal of a benchmark for the evaluation of imputation methods, motivated by a gap in the literature with respect to consistent and standardized protocols for the comparison between different approaches. A comprehensive set of 480 experiments was performed, with the evaluation of eight imputation methods, three public dataset, four missing patterns and five performance metrics through 5-fold