Jump to Search Jump to Navigation Jump to Content

Machine learning for lithology classification,data fusion and evaluating prediction uncertainty

CET Seminar Series

Matthew Cracknell

1). Machine learning algorithms (MLAs) are a group of powerful tools that have the ability to generate, with minimal user input, predictions for data-driven problems represented by a high-dimensional input space.
2). This research investigates the use of “off-the-shelf” machine learning algorithms (MLAs) for semi-automated supervised classification of geological phenomena.
3). A robust statistical comparison of several common MLAs has identified which are most suitable for predicting discrete geological units across a wide variety of terranes.