Innovative data mining approaches for outcome prediction of trauma patients

摘要: Trauma is the most common cause of death to young people and many these deaths are preventable [1]. The prediction trauma patients outcome was a difficult problem investigate till present times. In this study, models built their capabilities accurately predict mortality assessed. analysis includes comparison data mining techniques using classification, clustering association algorithms. Data were collected by Hellenic Emergency Surgery Society from 30 Greek hospitals. Dataset contains records 8544 suffering severe injuries year 2005 2006. Factors include patients' demographic elements several other variables registered time place accident until hospital treatment final outcome. Using obtained results compared in terms sensitivity, specificity, positive predictive value negative ROC curve depicts methods performance.