<![CDATA[Human liver belongs to one of the main body organs which functions for changing poisonous to be nutritious substances in order to control the hormone level of body. Consequently, liver disease can cause many problems in the human life. Fast and accurate prediction on it enables early and effective treatment. Nowadays, the advancement of computer technology has eased the diagnosis or decision-making process. Machine learning allows a computer to decide or give right suggestion. Trial on it has been carried out by fuzzy decision tree algorithm with some threshold values namely fuzziness control threshold (ðœƒð‘Ÿ) and leaf decision threshold (ðœƒð‘›) which were determined trial method of 10-fold cross validation and obtained accuracy by 78.95%. Since this accuracy was gained from distinct threshold values, the researcher used Receiver Operating Characteristics (ROC) method to investigate the performance of good accuracy. The result of ROC demonstrated that the most optimum accuracy by 78.95% was obtained on ðœƒð‘Ÿ=75%,77%,80%,82%,85%,87%,90%,92%,95%,98% and ðœƒð‘›=6%.Keywords: liver disease, Fuzzy Decision Tree, K-Fold Cross Validation, Receiver Operating Characteristics]]>
Copyrights © 2019
