Overview of some of the most important data science methods (statistics, machine learning & computer vision) relevant for geological sciences, as well as other fields in the Earth Sciences. Areas covered are: extreme value statistics for predicting rare events; compositional data analysis for geochemistry; multivariate analysis for designing data & computer experiments; probabilistic aggregation of evidence for spatial mapping; functional data analysis for multivariate environmental datasets, spatial regression and modeling spatial uncertainty with covariate information (geostatistics). Identification & learning of geo-objects with computer vision. Focus on practicality rather than theory. Matlab exercises on realistic data problems.
3 units · Letter or Credit/No Credit
Overview of some of the most important data science methods (statistics, machine learning & computer vision) relevant for geological sciences, as well as other fields in the Earth Sciences. Areas covered are: extreme value statistics for predicting rare events; compositional data analysis for geochemistry; multivariate analysis for designing data & computer experiments; probabilistic aggregation of evidence for spatial mapping; functional data analysis for multivariate environmental datasets, spatial regression and modeling spatial uncertainty with covariate information (geostatistics). Identification & learning of geo-objects with computer vision. Focus on practicality rather than theory. Matlab exercises on realistic data problems.
Offered in Winter 2027 at Stanford University.