A detailed Study of different Clustering Algorithms in Data Mining
Manish Gupta, Vikram Rajpoot, Ankur Chaturvedi, Ruchi Agrawal · 2022 2nd International Conference on Intelligent Technologies (CONIT) · 2022
Data mining (DM) is a practice in which large data stores are searched automatically to find designs as well as trends that go beyond simple analyses. Data mining is also known as the data clustering of knowledge discovery (KDD), where similar items are grouped into clusters. Clustering is one of the main analytical methods in DM. The clustering method directly influences clustering results. Clustering is the main as well as a key method for the automatic collection of data from enormous amounts. Its task is to recognize groups of indistinguishable objects in a data set, called clusters. Clustering methods, including database shopping, Web inspections, information acquisition, biotechnology, and more are used in a broad spectrum. This paper provides an overview of a few methods: partitioning, hierarchy, density-based method, Grid-based methodology, and data mining model-based method. Some of the clustering techniques have been compared among them based on the time to build a model, several instances, and several cluster formations. All of these algorithm's Expectation Minimization (EM) clustering algorithm takes more time and formed a maximum number of clusters in comparison to other clustering algorithms. But, the OPTICS algorithm is not better to cluster the number of instances.