Kateřina Koňasová - Measuring Dissimilarities Between Spatial Point Patterns
Kdy: 27/11/2019 14:00 - 27/11/2019 15:30
Kde: Seminární místnost MUUK (3. patro)
Spatial point processes are useful mathematical models that describe the arrangement of objects randomly placed in space. Such models are of particular interest in many scientific disciplines, including biology, ecology, particle physics or material science. In practical applications, a bunch of points on some bounded region in a plane or space is observed, whose number and locations are random. Such a collection of points is then called point pattern. Comparing two point patterns as well as describing in what manner they differ could be quite a demanding task. A tool for quantifying how similar or better dissimilar two point patterns are is thus needed.
We will see various examples of point pattern data and we will discuss different methods of measuring dissimilarities between point patterns. We especially aim to describe functional summary characteristics used in spatial statistics and outline how they can be applied to build a reasonable dissimilarity measure. To better illustrate the issue, these dissimilarity measures will be used to solve the classification task in the context of replicated point patterns.