Title :
Constrained Intelligent K-Means: Improving Results with Limited Previous Knowledge.
Author_Institution :
Birkbeck Coll., Univ. of London, London
fDate :
Sept. 29 2008-Oct. 4 2008
Abstract :
It is here presented a new method for clustering that uses very limited amount of labeled data, employees two pairwise rules, namely must link and cannot link and a single wise one, cannot cluster. It is demonstrated that the incorporation of these rules in the intelligent k-means algorithm may increase the accuracy of results, this is proven with experiments where the real number of clusters in the data is unknown to the method.
Keywords :
pattern classification; pattern clustering; clustering; constrained intelligent k-means algorithm; labeled data; Clustering algorithms; Computer applications; Data engineering; Data mining; Knowledge engineering; Partitioning algorithms; Semisupervised learning; Clustering; intelligent k-means; k-means; semi supervised clustering;
Conference_Titel :
Advanced Engineering Computing and Applications in Sciences, 2008. ADVCOMP '08. The Second International Conference on
Conference_Location :
Valencia
Print_ISBN :
978-0-7695-3369-8
Electronic_ISBN :
978-0-7695-3369-8
DOI :
10.1109/ADVCOMP.2008.30
