A Comparison of Missing Value Imputation Methods for Classifying Patient Outcome Following Trauma Injury

Kay Penny*, Thomas Chesney

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingPublished conference contribution

3 Citations (Scopus)

Abstract

A study is designed to compare several missing value imputation methods to enable classification of patient outcome following trauma injury. The Glasgow coma score is a measure of head injury severity, and is known to be important in determining patient outcome. The Glasgow coma scores are missing for 12% of the dataset, and in order to classify patient outcome for these patients, the missing values are first imputed. The first part of the study is designed to compare the performance of several missing value imputation methods, and errors between imputed values and known values of Glasgow coma scores are calculated. The second part of the study involves analysing the imputed data sets using logistic regression to classify whether patients live or die. Accuracy of results are compared in terms of sensitivity, specificity, positive predictive value and negative predictive value.
Original languageEnglish
Title of host publicationITI 2008 - 30th International Conference on Information Technology Interfaces
PublisherIEEE Explore
Pages367-370
DOIs
Publication statusPublished - 2008
EventInformation Technology Interfaces 2008 - Cavtat, Croatia
Duration: 23 Jun 200826 Jun 2008
Conference number: 30

Conference

ConferenceInformation Technology Interfaces 2008
Abbreviated titleITI
Country/TerritoryCroatia
CityCavtat
Period23/06/0826/06/08

Fingerprint

Dive into the research topics of 'A Comparison of Missing Value Imputation Methods for Classifying Patient Outcome Following Trauma Injury'. Together they form a unique fingerprint.

Cite this