Microclustering with Outlier Detection for DADC

Cluster analysis is a machine learning technique for categorizing unlabeled data. The data points are grouped into different clusters based on how similar they are. The objects that may be comparable are grouped together in a group with few or no similarities. Density based clustering algorithms, which can locate clusters of any shape while avoiding outliers, are used in many applications. Density based clustering algorithms consider dense sections of objects in the data space to be clusters, separated by low density areas (noise). The Domain Adaptive Density Clustering (DADC) technique was created to point out the issues of scattered cluster loss and cluster fragmentation. Micro clustering is a stream clustering technique that preserves compact data item information. Micro clusters estimate local density by combining data from several data points in a specific area. Micro-cluster is a time-based improvement to the cluster function that effectively compresses data. Incorrect data might appear in a database for a variety of reasons. Outlier identification is a technique for filtering irregularities generated in a database. In this work, we intend to put forward a method for micro clustering technique with outlier removal for Domain Adaptive Density Clustering.