Distributed and Memory Efficient Machine Learning for Spatial Analysis Applications

In the context of spatial analysis, spatial disaggregation or spatial downscaling are processes by which information at a coarse spatial scale is translated to finer scales, while maintaining consistency with the original dataset. Fine grained descriptions of geographical information is a key resource in fields such as social-economic studies, urban and regional planning, transport planning, or environment impact analysis. However, spatial disaggregation involves heavily iterative, and computational demanding algorithms, seriously compromising performance when the data sets at play involve large geospatial regions and contains large volumes of information. This paper presents a M.Sc. thesis project that concerns with the development of a scalable and efficient approach to spatial disaggregation, taking advantage of state-of-art approaches for parallel and distributed computation, as well as methods for leveraging