ABSTRACT
2.1 Introduction
The current popularity of data mining and data warehousing, as well as the decline in the cost of disk storage, has led to a proliferation of terabyte data warehouses [66]. Mining a database of even a few gigabytes is an arduous task for machine learning techniques and requires advanced parallel hardware and algorithms. An approach for dealing with the intractable problem of learning from huge databases is to select a small subset of data for learning [230]. Databases often contain redundant data. It would be convenient if large databases could be replaced by a small subset of representative patterns so that the accuracy of estimates (e.g., of probability density, dependencies, class boundaries) obtained from such a reduced set should be comparable to that obtained using the entire data set.
