On Cluster Validity for Fuzzy Clustering of Incomplete Data

Modified Possibilistic Fuzzy C-Means Algorithm for Clustering Incomplete Data Sets

“…In this study, centroid errors are the magnitude of the cluster centre error for an incomplete data set clustered using the OCSPFCM and NPSPFCM algorithms when compared to the cluster centre of a complete data set clustered using the PFCM. In some applications, knowing the cluster centres is important to determine the partitioning of data points [1]. Therefore, this research evaluates the two algorithms by calculating the centroid errors at each level of the missing values.…”

“…In some applications, information about the cluster center is important to know the data point partitioning in the cluster [1]. Therefore, we also evaluate the two algorithms by calculating the centroid errors at each level of the missing values percentage.…”

“…Incomplete data sets are commonly found in the real world due to failures during the collection, merging, cleaning, and transfer of data from one source to another [1]. The main problem faced when trying to cluster incomplete data sets is the inability of the existing algorithm to carry out the process.…”

“…Incomplete data sets are common in the real world. Incomplete data sets can be caused by failures during data collection, the problem of merging data from various data sources, data cleaning or data transfer problems [1]. The main problem faced when trying to cluster incomplete data sets is the existing clustering algorithm cannot be used for clustering incomplete data sets.…”