Training "Boosted Trees"-models and CNNs to predict facies type based on seismic data. Underlying Kaggle-competition: https://github.com/seg/2016-ml-contest.
For our "Digital prototyping"-class, we will be reenacting the results of the winners of the 2016 competition, and try to use the latest technologies (e.g. TensorFlow 2.0) to improve on the accuracy of the winner.