Title of Paper: A Prototype Clustering Technique for Data Mining and Knowledge Discovery
Principal Author: Ruixin Yang
Abstract: This presentation addresses a reliable,
feasible method to find geographical areas with constraints using hierarchical
depth-first clustering. The method involves multi-level hierarchical clustering
with depth-first strategy, depending on whether the area of each cluster is
satisfying given constraints. The attributes used in hierarchical clustering
are coordinates of grid data points. The constraints are an average value range
and the minimum size of area with a small proportion of missing data points.
Convex hull and point-in-polygon algorithms are involved in examining the
constraint satisfaction. The method is implemented for an Earth science data
set, NDVI, for vegetation studies.
This presentation also includes information about currently going development on web-based prototype of the above algorithm and implementation with relational database management system.