Machine Learning Algorithms for Analysis and Modeling of GeoSpatial Data

摘要

The paper presents an overview of the recently developed approaches and results concerning the application of Machine Learning algorithms for environmental data analysis and modelling problems. These include all the main tasks of supervised and unsupervised learning from spatio-temporal data. The incorporation of additional information into spatial models, such as secondary variables and/or co-ordinates, digital elevation models and CIS information is considered. The real case studies which are mentioned in the paper deal with the environmental problems of pollution data modelling and topo-climatic mapping.