New Graphical Symbolic Objects Representations in Parallel Coordinates

Abstract

Data visualization plays an outstanding role in descriptive statistics. Human eye has a strong ability in detecting regularities in the data and, in many cases, the analysis of graphed data can drive the analyst towards the choice of the most suitable analytical tools. Symbolic Data Analysis (SDA) aims at defining statistical methods to analyze complex data structures no longer based on the classical tabular model. In the SDA context, this paper proposes a thinking on the Symbolic Data visualization and, at the same time, new methods capable of representing complex data and preserving the statistical interpretation.