Abstract
How to extract relevant information from large data sets has become a main challenge in data visualization. Clustering techniques that classify data into groups according to similarity metrics are a suitable strategy to tackle this problem. Generally, these techniques are applied in the data space as an independent step previous to visualization. In this paper, we propose clustering on the perceptual space by maximizing the mutual information between the original data and the final visualization. With this purpose, we present a new information-theoretic framework based on the rate-distortion theory that allows us to achieve a maximally compressed data with a minimal signal distortion. Using this framework, we propose a methodology to design a visualization process that minimizes the information loss during the clustering process. Three application examples of the proposed methodology in different visualization techniques such as scatterplot, parallel coordinates, and summary trees are presented.
Highlights
Technology advances allow for obtaining large amounts of data related to any process in any application field
We have presented a new methodology to deal with the visualization of clustered data that ensures an optimal information transfer between the original data and the final visualization
We have presented a new mathematical framework, based on information theory and rate-distortion theory, that models the visualization as an information channel between the source data and the final user
Summary
Technology advances allow for obtaining large amounts of data related to any process in any application field. Examples include visual analysis of business data [1], scientific data [2], and images and videos [3], amongst others. Information visualization techniques have become a powerful tool to extract the valuable and useful information hidden in the data. A great variety of visualization techniques has been proposed [5], most of them lose their effectiveness when dealing with large data sets. Screen space limitations transform visualizations into cluttered images that are incomprehensible. To overcome these limitations data, clustering techniques can be applied
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have