Data mining is an investigational discipline and its main goal is to get an accurate prediction of any assigned task. Feature selection is highly relevant in predictive analysis and should not be ignored, it helps reduce the execution time and provides a more accurate and reliable result. Many studies have been conducted in this field, but more research on predictive analysis and how reliable these predictions should be is still required. Application of data mining techniques in the health sector ensures that the right diagnosis, treatment and positive result are given to patients. This study was implemented using the WEKA tool. The study was designed using three classifiers (Naïve Baye, J48 Decision Tree and Support Vector Machine) for the prediction of the tropical diabetes hand syndrome dataset. The performance of the classifiers was evaluated considering their Accuracy, Specificity, Sensitivity, Error rate and Precision. Based on the performance metrics, results showed that Naïve baye gives the best result and the Support Vector Machine (SVM) had the lowest execution time, making it the fastest classifier.
Key words: Diabetes hand Syndrome, Prediction, Naïve bayes, J48 Decision tree, Support Vector Machine (SVM).
|
