Using Projection Techniques on the Exploratory Analysis of Multidimensional Data
Fernando Vieira Paulovich
Univ. de Sao Paulo, Sao Carlos/SP (Brazil)

Abstract:

The analysis of multidimensional (large) data sets remains as challenge for many areas of computer sciences and statistics. These features are present in many types of data sets, going from document collections to data captured by sensors or generated on simulations. Amongst the various information visualization techniques which claim to be effective on supporting the interpretation of such data, distance-based multidimensional projection, also known as multidimensional scaling (MDS), has shown good results on helping users to extract patterns based on similarity. In this colloquium, the idea of projection techniques is introduced, presenting how they can be applied on exploratory and discovery tasks involving text and music collections, biosensors datasets and multivariate simulations.

Biography:

Fernando Vieira Paulovich obtained his bachelor and master degree in Computer Science at Universidade Federal de São Carlos, São Carlos/SP (Brazil) in 2000 and 2003, respectively. He received his Ph.D. in Computer Science in 2008 at Universidade de São Paulo, São Carlos/SP (Brazil), with one year internship at Delft University of Technology (TU-Delft), Deft (The Netherlands). His main fields of interest are information visualization, visual data mining and visual analytics. Currently, he is a lecturer and researcher at Universidade de São Paulo, São Carlos/SP (Brazil).

back to the list of talks