Quality-Based Visualization Matrices

Parallel coordinates and scatterplot matrices are widely used to visualize multi-dimensional data sets. But these visualization techniques are insufficient when the number of dimensions grows. To solve this problem, different approaches to preselect the best views or dimensions have been proposed in the last years. However, there are still several shortcomings to these methods. In this paper we present three new methods to explore multivariate data sets: a parallel coordinates matrix, in analogy to the well-known scatterplot matrix, a classbased scatterplot matrix that aims at finding good projections for each class pair, and an importance aware algorithm to sort the dimensions of scatterplot and parallel coordinates matrices.