Validation of Clustering Based Framework Using Unsupervised Machine Learning

K-Means algorithm of clustering is reflected in a more influential and widespread concept of mining in the research community. But, in spite of its acceptance, this method has few limitations associated with initialization problem of centroids to lead to surprise converge. Therefore, it is responsible for various shapes of the cluster and also affects the outlier. But the main problem of its method is that it is not capable to manage the huge dataset. In this paper, we have discussed how to overcome the limitations, where their performance is a measure based on experimental analysis by using the iris data set. In this paper, we proposed an efficient framework data mining using machine learning unsupervised technique k-means, this framework system which is implemented with MATLAB software, and NCSS 2021, SPSS tools for simulation to assess the training of iris, and heart disease datasets to measure the validity of clusters. Furthermore, proposed concept leads to the various directions of studies and exploration for research.