Comparison of Knowledge-Based Reasoning Methods to Measure the Effectiveness of Diagnostic Results
Open Access
- 1 February 2021
- journal article
- research article
- Published by IOP Publishing in Journal of Physics: Conference Series
- Vol. 1783 (1), 012049
- https://doi.org/10.1088/1742-6596/1783/1/012049
Abstract
This study describes a comparative analysis of methods in knowledge-based reasoning which aims to determine the best and optimal method in producing a diagnosis of a disease. The methods to be compared are Bayes' Theorem, Certainty Factor, and Euclidean Probability. The three methods were chosen because they can generate hypotheses from several existing possibilities, this can be seen from the many uses of these methods in previous studies. With this research, it can be used as material for consideration or support in producing a diagnosis conclusion. The process of testing the comparison of these methods is carried out by selecting the highest diagnosis result value and performing a comparative analysis using exponential techniques. The results of this test indicate that the Euclidean Probability produces an accuracy value of 75%, then the Bayes Theorem obtains an accuracy value of 62% and the Certainty Factor obtains an accuracy value of 87%. In addition, the results of the comparison of methods using exponential techniques show that Euclidean Probability gets 84%, then Bayes' Theorem gets 78% and the Certainty Factor gets 87%. With these results, it can be concluded that the Certainty Factor is better than Euclidean Probability and Bayes' Theorem in diagnosing disease.Keywords
This publication has 3 references indexed in Scilit:
- A feature dependent Naive Bayes approach and its application to the software defect prediction problemApplied Soft Computing, 2017
- Aplikasi Sistem Pakar Diagnosa Penyakit Pada Anak Bawah Lima Tahun Menggunakan Metode Forward ChainingJournal of Information Systems Engineering and Business Intelligence, 2017
- SODE: Self-Adaptive One-Dependence Estimators for classificationPattern Recognition, 2016