Export to Refworks Export to EndNote Email
Report an Issue

Determining the Number of Clusters in a Data Set Without Graphical Interpretation

Author:
Aguirre, Nathan S.
Published:
August 06, 2011.
Physical Description:
1 electronic document
Additional Creators:
Davies, Misty D.
Online Version

Availability

I Want It

Finding items...

Restrictions on Access:
Unclassified, Unlimited, Publicly available.
Free-to-read Unrestricted online access
Summary:
Cluster analysis is a data mining technique that is meant ot simplify the process of classifying data points. The basic clustering process requires an input of data points and the number of clusters wanted. The clustering algorithm will then pick starting C points for the clusters, which can be either random spatial points or random data points. It then assigns each data point to the nearest C point where "nearest usually means Euclidean distance, but some algorithms use another criterion. The next step is determining whether the clustering arrangement this found is within a certain tolerance. If it falls within this tolerance, the process ends. Otherwise the C points are adjusted based on how many data points are in each cluster, and the steps repeat until the algorithm converges,
Other Subject(s):
Collection:
NASA Technical Reports Server (NTRS) Collection.
Note:
Document ID: 20110016534.
ARC-E-DAA-TN3997.
Terms of Use and Reproduction:
Copyright, Distribution as joint owner in the copyright.
View MARC record | catkey: 15982154