Big Geospatial Data

Big Geospatial Data

Dozent: Prof. Philipp Otto

Qualification Goals

The module (2V/2Ü) conveys principles and methods of advanced geo data analysis and processing. The students are acquainted with the state of the art in spatial data anylsis and they get in contact with current research examples and the mathematical, theoretical background of the models. Further, they learn methods and infrastructures for parallel computing with very large datasets and of methods for parallel processing of geospatial data. After successful participation, they are able to assess and independently employ suitable frameworks and approaches for project realizations.

Module Contents

First, basics in spatial data analysis and mining are discussed along with methods to draw conclusions from data. Further, fundamentals of parallel computing are discussed, that is, when algorithms can be processed in parallel, computational complexity, and methods of parallel computing. Following this, established approaches to process spatial data are covered. For that, aggregation functions (e.g. mean values, local entropy, rasterization, hotspot detection), data locality, statistical testing, and further topics are discussed based on examples.