WeVoS-ViSOM: An ensemble summarization algorithm for enhanced data visualization

This study presents a novel version of the Visualization Induced Self-Organizing Map based on the application of a new fusion algorithm for summarizing the results of an ensemble of topology preserving mapping models. This algorithm is referred to as Weighted Voting Superposition (WeVoS). Its main feature is the preservation of the topology of the map, in order to obtain the most truthful visualization of datasets under study as possible. To achieve this, a weighted voting process takes place between the units of the maps in the ensemble in order to determine the characteristics of the units of the resulting map. In order to present a thorough study of its capabilities, several di erent quality measures have been applied and analysed under this novel neural architecture called WeVoS-ViSOM. To complete the study,it has also been compared with with the well-know SOM and its fusion version, the WeVoS-SOM and with two other previously devised fusion algorithms Fusion by Euclidean Distance and Fusion by Voronoi Polygon Similarity based on the analysis of the previous same quality measures in order to present a thorough study of its capabilities. All three summarization methods were applied to three widely used datasets from the UCI Repository and after a rigorous performance analysis, it is clearly demonstrated that the novel fusion algorithm outperformed the other single and summarization methods in terms of visualization of the datasets.