XVDM is a novel system for visual data mining, which enables joined comparison, investigation and learning of different techniques. The existing approaches focus mostly on the visualization of the results: they are not applicable for the investigation of several techniques at once and are not efficient for comparison by slight change in the dataset or input parameters. The salient properties of the XVDM are: it enables construction of synthetics datasets in a real time; it supports unlimited number of data mining techniques to be run at the same time; it provides an intuitive interface to switch between different techniques and to compare the results of them numerically or visually.
The GTree is the core data struucture that is implemented inside the XVDM system. The GTree is used to efficiently compute and query multidimensional density information.