A comparison of data mining methods and logistic regression to determine factors associated with death following injury.

Penny, Kay I; Chesney, Thomas

Abstract

A comparison of techniques for analysing trauma injury data collected over ten years at a hospital trauma unit in the U.K. is reported. The analysis includes a comparison of four data mining techniques to determine factors associated with death following injury. The techniques include a classification and regression tree algorithm, a classification algorithm, a neural network and logistic regression. As well as techniques within the data mining framework, conventional logistic regression modelling is also included for comparison. Results are compared in terms of sensitivity, specificity, positive predictive value and negative predictive value.

Citation

Penny, K. I., & Chesney, T. (2006). A comparison of data mining methods and logistic regression to determine factors associated with death following injury. In Data analysis, classification and the forward search: proceedings of the meeting of the Classification and Data Analysis Group (CLADAG) of the Italian Statistical Society, Univeristy of Parma June 6-8 2005 (417-423). Springer Berlin Heidelberg. https://doi.org/10.1007/3-540-35978-8_46

• Publication Date: Aug 6, 2006
• Deposit Date: Apr 4, 2008
• Peer Reviewed: Peer Reviewed
• Pages: 417-423
• Book Title: Data analysis, classification and the forward search: proceedings of the meeting of the Classification and Data Analysis Group (CLADAG) of the Italian Statistical Society, Univeristy of Parma June 6-8 2005
• ISBN: 9783540359777
• DOI: https://doi.org/10.1007/3-540-35978-8_46
• Keywords: Trauma injury; Data mining; Classification and regression tree algorithm; Classification algorithm; Neural network; Logistic regression
• Public URL: http://researchrepository.napier.ac.uk/id/eprint/1578
• Publisher URL: http://dx.doi.org/10.1007/3-540-35978-8_46