Visualization with Mining in the Context of Data Science
Dr Rosane Minghim, Associate Professor, University of São Paulo, Brasilien.
Abstract
Visual Analytics is a research field of great interest that has progressed considerably in the last years. It relies heavily in both mining and visualization strategies and is an important component of the data science discipline. In this seminar, we intend to stimulate the discussion on the ways that mining tasks and tools can benefit from visualization techniques, and to illustrate how mining strategies can enhance visualizations.
Coming from the viewpoint of visualization research, a few techniques will be mentioned (such as dimension reduction, multidimensional projections, tree-based visualizations), and we will present some of our previous work on developing and employing these techniques to generate visualizations and to support mining tasks. We hope to finish the seminar with a discussion on ways that visualization can move forward as a supporting tool for data mining of increasingly complex, and on ways to continue research on integration between these two fields.
Bio
Rosane Minghim is an Associate Professor in Visual Analytics at University of São Paulo (USP), Brazil. Her fields of interest include all issues related to Data Visualization and Visual Analytics, as well as applications. Her background includes undergraduate and postgraduate work at USP, University of Campinas, and University of East Anglia (UK), as well as periods as a visiting researcher at both the Faculty of Computer Science, Dalhousie University, Halifax, Canada (2015) and University of Massachussetts Lowell (2001/2002).