DUGi

Handling Missing Phenotype Data with Random Forests for Diabetes Risk Prognosis

http://dugi.udg.edu/item/http:@@@@hdl.handle.net@@2072@@319217

European Conference on Artificial Intelligence (ECAI), The Hage, Netherlands, 29 August - 2 September 2016 (Session 1st ECAI Workshop on Artificial Intelligence for Diabetes)

Machine learning techniques are the cornerstone to handle the amounts of information available for building comprehensive models for decision support in medical practice. However, the datasets use to have a lot of missing information. In this work we analyse how the random forests technique could be used for dealing with missing phenotype values in order to prognosticate diabetes type 2

This project has received funding from the grant of the University of Girona 2016-2018 (MPCUdG2016) and the European Unions Horizon 2020 research and innovation programme under grant agreement No 689810 (PEPPER). The work has been developed with the support of the research group SITES awarded with distinction by the Generalitat de Catalunya (SGR 2014-2016)

European Conference on Artificial Intelligence (ECAI)