An event series for Turing-Roche partnership updates, knowledge sharing and new perspectives.

Speaker: Dr Robin Mitra, Professor in Statistics at University College London

Our capacity to store and analyse ever-larger data sets offers many potential benefits in developing better methodology for dealing with important research challenges, such as building predictive models that can take account of underlying heterogeneity in the population. However, an increasingly encountered challenge to overcome in this setting is the presence of Structured Missingness (SM) in the data. This is where missing values are present in the data and also exhibit an underlying association or structure.

SM may often, but not exclusively, manifest as missing data blocks. A common instance where SM might arise is when data are combined at scale across different modalities, for example in healthcare settings when seeking to combine electronic health records with genomic information and imaging data. Investigating the problem of SM has thus been a key theme of the Turing-Roche partnership. In this seminar, we review the research activities into SM undertaken through the partnership as well as describe our current areas of investigation.

Find out more and register at: Turing-Roche Knowledge Share Series: Structured missingness - a review and current research directions | The Alan Turing Institute