Research Article
Effect of Fuzzy Criteria on the Performance of Decision Tree Models for Heart Disease Classification
@INPROCEEDINGS{10.4108/eai.3-11-2023.2347907, author={Endang Sri Kresnawati and Des Alwine Zayanti and Ali Amran and Yulia Resti and Nada Aisyah and Anggraini Salsabila}, title={Effect of Fuzzy Criteria on the Performance of Decision Tree Models for Heart Disease Classification}, proceedings={Proceedings of the 3rd Sriwijaya International Conference on Basic and Applied Sciences, SICBAS 2023, November 3, 2023, Palembang, Indonesia}, publisher={EAI}, proceedings_a={SICBAS}, year={2024}, month={8}, keywords={decision tree fuzzy criteria accuracy precision recall}, doi={10.4108/eai.3-11-2023.2347907} }- Endang Sri Kresnawati
Des Alwine Zayanti
Ali Amran
Yulia Resti
Nada Aisyah
Anggraini Salsabila
Year: 2024
Effect of Fuzzy Criteria on the Performance of Decision Tree Models for Heart Disease Classification
SICBAS
EAI
DOI: 10.4108/eai.3-11-2023.2347907
Abstract
Fuzzy is a special algorithm used in developing Decision Tree Models. The decision rules obtained not only depend on the tree structure formed, but also on the determina-tion of numerical criteria and fuzzy linguistic criteria. This study aims to measure the performance of decision tree models in diagnosis heart disease patients. Models were built using two, three, four, and five fuzzy criteria. The research stages start from dis-cretization, building a decision tree structure, developing decision rules, and evaluating model performance. The calculation results show that model performance increases in models using two and three fuzzy criteria, then decreases in models with four and five criteria. The model with three criteria provides better performance. The best perfor-mance was obtained from a model with three criteria, namely 73.33% accuracy, 77.08% precision and 74% recall.