Title :
A New Clustering Algorithm Compared With the Simple K-Means
Author_Institution :
Inst. of Telecommun., Xi´´an Jiaotong Univ., Xi´´an, China
Abstract :
This paper presents a new clustering technique named Chinese Reomi´s Algorithm (CR for short), for we´re Chinese and Reomi is an English name. Because the CR is very much like the simple k-means mechanism, we take the two for comparison. Further more, we suggest a simple but useful way to estimate the number of clusters.
Keywords :
pattern clustering; Chinese algorithm; Reomi algorithm; clustering algorithm; k-means algorithm; Chromium; Clustering algorithms; Data mining; Information analysis; Information retrieval; Iterative algorithms; Partitioning algorithms; Pattern analysis; Pattern recognition; Performance gain;
Conference_Titel :
Management and Service Science, 2009. MASS '09. International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-4638-4
Electronic_ISBN :
978-1-4244-4639-1
DOI :
10.1109/ICMSS.2009.5302386
