
126 Big Data: Techniques and Technologies in Geoinformatics
that include only nearby points that can have an effect. An example of this is used
often in the remote sensing community where a subset of pixels around a given pixel
isincluded in the feature vector for modeling the center pixel as shown in Figure 6.2.
Big data volumes allow ltering to be done more effectively since we will always
have sufcient data in our neighborhood to model with. Juxtapose this to traditional
data accumulation that often left the modeler with a sparse sample of the underlying
phenomenon and required algorithms that could maximize the information gained
across ...