Recently, various clustering mechanism has been introduced to form data and cluster them into diverse domains. Some of the clustering algorithms clustered the data in proper way for grouping datasets accurately. However, some of the clustering methods roughly merge the categorized and numeric data types. Clustering is a process to identify the patterns distribution and intrinsic correlations in large datasets by separation of data points into similar classes. The proposed system, Improved Linked Clustering (ILC), is introduced to find a number of clusters on mixed datasets to produce results for several datasets. The ILC algorithm helps to prefer which clustering mechanism should be utilized to obtain a coherent and a mixed data significant mechanism for specific character deployment. Moreover, the technique can be used for optimization criteria during cluster formation for assisting clustering process toward better and efficient character interpretable technique. The technique objective is to offer a novel clustering method for data clustering methods evaluation over mixture of datasets including prior spatial information about the relation of elements which represents the clusters. The method provides an idea to estimate the character significance of clustering technique. The method estimates the summarization of the spatial point analysis of clustering technique with respect to coherence and clusters distribution. The proposed method is evaluated in two databases (character, spatial) using three conventional clustering techniques. Based on experimental evaluations, the proposed ILC algorithm improves the 0.04% cluster accuracy and 0.4 s cluster computation time compared to conventional techniques on the spatial dataset.
CITATION STYLE
Lakshmaiah, K., Krishna, S. M., & Reddy, B. E. (2020). Design an improved linked clustering algorithm for spatial data mining. In Advances in Intelligent Systems and Computing (Vol. 1090, pp. 403–416). Springer. https://doi.org/10.1007/978-981-15-1480-7_34
Mendeley helps you to discover research relevant for your work.