Machine Learning of Environmental/Geospatial Data
Mikhail Kanevski, Vasily Demyanov
This workshop will cover geospatial data machine learning: concepts, algorithms and case studies.
The workshop aims to demonstrate the value of machine learning in solving challenging geospatial prediction problems. The workshop will introduce concepts and good practices of data driven based modelling and provide an overview of some commonly used algorithms through a series of multi-disciplinary case studies in complex spatial data modelling: environmental data mining and mapping (pollution, natural hazards, and renewable resources), prediction of subsurface geological systems behaviour. The workshop will demonstrate how contemporary machine learning approach can help to overcome limitations of traditional geostatistics. A particular emphasis will be made on handling large volumes of spatial/temporal data and dealing with uncertainty and model predictions. The workshop aims to attract interest across multidisciplinary audience of academics and practitioners, who deal with large amounts of spatial/temporal multivariate information.
The workshop will include several keynote overview and application presentations as well as pitch presentations to contribute open problems and cases to the workshop discussion.
