Fuzzy Rule Based Classification for Heart Dataset using Fuzzy Decision Tree Algorithm based on Fuzzy RDBMS
Idris Mala, Pervez Akhtar, Tariq Javid, Syed Saood Zia, Sir Syed · 2013
4 Abstract: A fuzzy rule-based system design concentrates on accuracy and interpretability of the system. Fuzzy decision tree method is proposed based on fuzzy RDBMS and rule generation based on C4.5 algorithm known as fuzzy rule generation system (FRGS) algorithm. A fuzzy decision tree is developed by first converting a medical application of heart relational database to fuzzy heart relational database and then developing the decision tree using C4.5 algorithm. In the next step rule generation is performed from the C4.5 algorithm. A pruning process is performed to prevent overfitting. Different pruning rates are analyzed to show the variation of interpretability in the generated model. Results show that good tradeoff between accuracy and interpretability can be made by varying pruning.