Title :
Clustering Algorithms Based on Mahalanobis Distances
Author :
Yih, Jeng-Ming ; Lin, Yuan-Horng
Author_Institution :
Dept. of Math. Educ., Nat. Taichung Univ., Taichung, Taiwan
Abstract :
Fuzzy c-means algorithm (FCM) based on Euclidean distance function converges to a local minimum of the objective function, which can only be used to detect spherical structural clusters. In this paper, an improved Fuzzy C-Means algorithm based on a Normalized Mahalanobis distance by taking a new threshold value and a new convergent process is proposed. The experimental results of three real data sets containing image classification show that our proposed new algorithm has the better performance.
Keywords :
fuzzy set theory; pattern classification; pattern clustering; Euclidean distance function; FCM; Mahalanobis distances; clustering algorithms; fuzzy c-means algorithm; image classification; spherical structural; threshold value; Accuracy; Algorithm design and analysis; Classification algorithms; Clustering algorithms; Equations; Euclidean distance; Iris; Clustering Algorithms; FCM; Mahalanobis Distances;
Conference_Titel :
Electronic Commerce and Security (ISECS), 2010 Third International Symposium on
Conference_Location :
Guangzhou
Print_ISBN :
978-1-4244-8231-3
Electronic_ISBN :
978-1-4244-8231-3
DOI :
10.1109/ISECS.2010.57
