Spatial Subgroup Mining Integrated in an Object-Relational Spatial Database
PKDD '02 Proceedings of the 6th European Conference on Principles of Data Mining and Knowledge Discovery
Mining census data for spatial effects on mortality
Intelligent Data Analysis
ECML PKDD '08 Proceedings of the European conference on Machine Learning and Knowledge Discovery in Databases - Part II
Subgroup mining for interactive knowledge refinement
AIME'05 Proceedings of the 10th conference on Artificial Intelligence in Medicine
Hi-index | 0.00 |
Micro data are often available for several time points, especially when new data are incrementally collected, for instance by regularly adding new batches of objects (e.g., daily or monthly). In this article, we summarize subgroup mining approaches to analyze several cross-sections of data, each representing a special time point. We assume the more general case of independent cross-sections not necessarily containing the same objects. Change patterns are then typically more useful for an analyst, since in this regularly proceeding or incremental situation, the main static patterns related to a special time point (see Chapter 16.3.1 of this handbook) are often quite stable over time and mostly well known. Specifically we deal with analyzing change (two time points) or trend (sequence of equidistant time points), and we discuss pattern elaboration to refine or combine diverse types of patterns and to deal with some pecularities (Simpson's paradox).