Criterion in selecting the clustering algorithm in radial basis functional link nets
Ang Sau Loong, Ong Hong Choon, Low Heng Chin · WSEAS TRANSACTIONS on SYSTEMS archive · 2008
K-means and k-median clustering algorithms can help in the selection of centres for the Radial Basis Functional Link Nets. Radial Basis Functional Link Nets is used to classify the data. In this paper, we will show the importance of knowing the skewness of the data in deciding to choose between k-means or k-median clustering algorithm in finding the centre of Radial Basis Functional Link Nets and we will also show that this initial selection criterion will result in the improvement of efficiency in terms of speed and accuracy in data classification. Two sets of real data are used to demonstrate our results.