Excavating clusters of star coordinate visualization through dimension manipulation / Izyan Izzati Kamsani

Recent advancement in high-performance technologies has resulted in collections of high dimensional data. High dimensional data consist of large number of data records and attributes. It is challenging when it comes to extracting high dimensional data into meaningful information. Cluster analysis is used as a method to provide useful summary so that it can be understood more straightforwardly. The summary obtained, however is difficult to interpret from raw data. In order to understand the clusters results, it is visualized in a low dimensional plane. There are many available visualization techniques that can handle high dimensional data. One of it is geometric projection technique. Examples of geometric projection techniques are; Scatterplot (SP), Parallel Coordinate (PC) and Star Coordinate (SC) visualization techniques. However, for this study, only SC visualization technique was explored in further details. SC can display high dimensional data into a limited space. SC performs by plotting data dimensions in a circular arrangement.