next up previous
Next: 2 Software Up: Database Internals Software Project Previous: Database Internals Software Project

1 Introduction

As many people are aware there is a growing need for new methods in the area of data mining. One such area in which there are huge amounts of data which need to be processed is GIS, or geographical information systems.

To this end we decided to look at one of the ways to get data associated with other pieces of data, called clustering. Clustering can have an extremely bad time complexity for naive algorithms. For this reason we have looked at new ways of clustering such data. We found a new clustering algorithm for \( k\)-MEDOIDS which has a sub-quadratic time complexity.

Kevin Pulo