An efficient ICKM approach for similarity measurement and distance estimation based on k-means

Abstract

An iterative centroid initialization k-means (ICKM) based clustering has been proposed in this paper. In this approach first the dataset selection has been performed along with the option of choosing and selection as per the data use or the user can access partial data also based on the iterative centroid.Then the data preprocessing steps are followed for the data arrangement and analysis. There are four different distance algorithms have been considered with the k-means. These algorithms provide the complete variability for the distance estimation and production. The proposed method found to be useful along with different distance estimation and measures.

Analysis of interconnected configuration for penetration of distributed generation

“…Computational Analysis via multiple machine learning algorithms will be performed in the future to find out probable values of case-specific voltage stability index [20]. K-Means clustering [21], Support Vector Machine (SVM), decision tree, random forest, and multilayer perception (MLP) algorithms [22] will be used in the future for creating clusters of similar and different VSI values for comparative studies. Finally, the incorporation of innovation in the power industry with the addition of Information and Communication Technologies (ICT) in terms of sustainable development will have a great impact on power distribution management [23].…”