Prediction of different types of liver diseases using rule based classification model

摘要: BACKGROUND: Diagnosing different types of liver diseases clinically is a quite hectic process because patients have to undergo large numbers independent laboratory tests. On the basis results and analysis test, are classified. Hence simplify this complex process, we developed Rule Base Classification Model (RBCM) predict diseases. The proposed model combination rules data mining techniques.OBJECTIVE: objective paper propose rule based classification with machine learning techniques for prediction Liver diseases.METHOD: A dataset was twelve attributes that include records 583 in which 441 were male rests female. Support Vector Machine (SVM), Induction (RI), Decision Tree (DT), Naive Bayes (NB) Artificial Neural Network (ANN) K-cross fold technique used performance these evaluated accuracy, sensitivity, specificity kappa parameters as well statistical (ANOVA Chi square test) analyze disease independence attributes.RESULT: Out patients, 416 affected 167 healthy. decision tree (DT) provides better result among all (RI, SVM, ANN NB) (Accuracy 98.46%, Sensitivity 95.7%, Specificity 95.28% Kappa 0.983) while SVM exhibits poor 82.33%, 68.03%, 91.28% 0.801). It also found best without Accuracy 82.68%, 86.34%, 90.51% 0.619) almost similar worst (SVM, 0.801 accuracy chi test 76.67%.CONCLUSION: This study demonstrates there significant difference between most accurate result. can be valuable tool medical making.