Description: This paper studies the problem of categori cal data clustering. especially for transactional data characteri propellant by high dimensionality and large volume. St. arting from a heuristic method of increasing th e height-to-width ratio of the cluster histogr am, we develop a novel algorithm-CLOPE. which is very fast and scalable, while being quite effective. We demonstrate th e performance of our algorithm on two real world
- [noseecluster] - clustering analysis techniques have wide
- [pmt2007] - there are some other approaches. such as
- [BEAlgorithm2007] - Based on the data mining algorithms to a
- [K_Means] - k-means is a classical clustering algori
- [MFPBCLU] - I wrote a paper on the GML-based cluster
File list (Check if you may need any files):
Download