KDnuggets : News : 2004 : n07 : item16 | previous | next |
SoftwareFrom: Kamal KarlapalemDate: Sat, 3 Apr 2004 Subject: ReCkless Data Clustering Algorithms ReCkless is a suite of clustering algorithms based on the concept of k-Reverse Nearest Neighbors developed at the Centre for Data Engineering. It performs outlier detection to eliminate noise/outliers before executing cluster identification. You can test the ReCkless aglomerative clustering algorithm by giving the data set through web, and get back the result. It currently works for multi-dimensional numerical data sets. The url is: http://cde.iiit.net/RNNs/ Users need to register at the site to run the algorithm over the datasets. Any questions, comments, problems and bugs can be addressed to reckless@students.iiit.net
|
KDnuggets : News : 2004 : n07 : item16 | previous | next |
Copyright © 2004 KDnuggets. Subscribe to KDnuggets News!