Open AccessThe Efficient Distance Weighted Case Base Rule (DW-CBR) for Early Childhood Diseases Diagnosis
Author(s) -
Indah Werdiningsih,
Rimuljo Hendradi,
P. Purbandini,
Barry Nuqoba,
Elly Anna
Publication year - 2021
Publication title -
computing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.184
H-Index - 11
eISSN - 2312-5381
pISSN - 1727-6209
DOI - 10.47839/ijc.20.2.2174
Subject(s) - case based reasoning , similarity (geometry) , model based reasoning , computer science , artificial intelligence , rule based system , reasoning system , knowledge base , qualitative reasoning , opportunistic reasoning , expert system , data mining , machine learning , knowledge representation and reasoning , image (mathematics)
Children from newborns to six years old are more susceptible to diseases. A common methodology to diagnose childhood diseases is by using a reasoning technique. Reasoning techniques is one of a reliable method for expert systems. Reasoning techniques using the correct case of results have provided enormous support for predicting the diagnosis and treatment of diseases. This paper focuses on the main technical characteristics of two common reasoning techniques, namely; rule-based reasoning and case-based reasoning. This paper describes a comparative analysis of rule-based and case-based reasoning techniques using several commonly used similarity measures and a study on its performance for classification tasks. Moreover, this study proposes a new case-based reasoning approach using an alternative similarity measure, called Distance-Weighted Case Base Reasoning (DW-CBR). The proposed method aims to improve classification performance. The main result of this study shows that case-based reasoning is a more powerful methodology regarding the issues of maintenance and knowledge representations over the rule-based system and reveals that DWCBR has the best accuracy, which is 92%.