School of Computing Research Colloquia

To See What Isn’t There – Visual Analysis of Missing Data

Department of Computer Science and Digital Technologies, Northumbria University

Abstract: Missing data are records that are absent from a data set. They are data that were intended to be recorded, but for some reason were not. Missing data occur in almost any domain and is a common data analysis challenge that causes problems such as biased results and reduced statistical rigour. Although visualization has great potential to provide invaluable support for the investigation of missing data, missing data challenges are rarely addressed by the visualization society. This talk will cover various concepts and aspects in missing data analysis, suggesting patterns of relevance for gaining further understanding of ‘missingness’ in datasets and present visualization approaches that have the potential to support this understanding