Dendrograms-based disclosure method for evaluating cluster analysis in the IoT domain

The Internet of Things (IoT) generates huge amount of data at an extremely fast pace. Thus, it is important to classify such data objects into different groups or clusters in order to gain some valuable insights from data. This paper aims to develop a dendrograms-based method for 3D visualization of hierarchical clustering for multidimensional data which can be collected from IoT devices and open databases. This method is built on hierarchical clustering algorithm which is simple and efficient. It presents areas of the selected clusters and their objects on a plane, according to the coordinates defined by the open dendrogram. It defines rules for visualization of the dendrogram and allows to find the nature of clusters. The paper also proposes quantitative indicators of localization of objects and evaluation of clusters being formed. The proposed method is evaluated using IoT-based dataset prepared in two different forms. The proposed method significantly improves the quality of visualization and evaluation of cluster analysis results. It is also efficient as the time complexity is significantly less for factorial analysis.

Automated summary

This paper presents a dendrograms-based method for 3D visualization of hierarchical clustering of multidimensional data collected from IoT devices and open databases. The method, built on a simple and efficient hierarchical clustering algorithm, significantly improves the quality of visualization and evaluation of cluster analysis results through rule definitions and quantitative indicators.