Extreme learning machine (ELM) is an emerging, and efficient learning algorithm for the single hidden layer feed forward neural network (SLFN). ELM and its variants are mainly used for classification and regression problems. Recently few research studies address the capability of extreme learning machine in the clustering application. In this work we proposed a novel clustering approach which combines the ELM and k-Means++ techniques for clustering. ELM was used to project the data items into

more »

... e higher dimensional space and k-Mean++ techniques used to perform the clustering in this feature space. The proposed method was tested on four different dataset and compared with three different algorithm like k-Means, K-Means++ and ELM k-Means++. The obtained result shows that the novel ELM with k-Means++ approach gives the better accuracy compared with other three clustering techniques.