Title :
Predicting Burn Patient Survivability Using Decision Tree In WEKA Environment
Author :
Patil, B.M. ; Toshniwal, Durga ; Joshi, R.C.
Author_Institution :
Dept. of Electron. & Comput. Eng., Indian Inst. of Technol. Roorkee, Roorkee
Abstract :
The use of data mining approaches in the domain of medicine is increasing rapidly. The effectiveness of these approaches to classification and prediction has improved the performance of their systems. These are particularly useful to medical practitioners in decision making. In this paper, we present an analysis of prediction of the survivability of the burn patients. The machine learning algorithm c4.5 is used to classify the patients using WEKA tool. The performance of the algorithm is examined by using the classification accuracy, sensitivity, specificity and confusion matrix. The dataset was collected from Swami Ramanand Tirth Hospital, Ambajogai, Maharashtra, India and is used retroactively from data records of the burn patients. The results are found to be precise and accurate by comparing with actual information on survivability or death.
Keywords :
data mining; decision trees; learning (artificial intelligence); medical computing; prediction theory; WEKA environment; WEKA tool; burn patient survivability prediction; classification accuracy; confusion matrix; data mining; decision tree; machine learning; medicine; Algorithm design and analysis; Biomedical engineering; Classification algorithms; Data analysis; Data engineering; Data mining; Decision trees; Diseases; Machine learning algorithms; Predictive models; Burn Patient; Data Mining; Prediction; WEKA;
Conference_Titel :
Advance Computing Conference, 2009. IACC 2009. IEEE International
Conference_Location :
Patiala
Print_ISBN :
978-1-4244-2927-1
Electronic_ISBN :
978-1-4244-2928-8
DOI :
10.1109/IADCC.2009.4809213
