IMPLEMENTATION OF CASE BASED REASONING FOR DIAGNOSING TUBERCULOSIS DISEASE USING K-NEAREST NEIGHBOR

J-Icon : Jurnal Komputer dan Informatika

View Publication Info
 
 
Field Value
 
Title IMPLEMENTATION OF CASE BASED REASONING FOR DIAGNOSING TUBERCULOSIS DISEASE USING K-NEAREST NEIGHBOR
IMPLEMENTASI CASE BASED REASONINGUNTUK MENDIAGNOSIS PENYAKIT TUBERKULOSIS MENGGUNAKANALGORITMA K-NEAREST NEIGHBOR
 
Creator Atok, Emanuel Tes
Sina, Derwin R
Sihotang, Dony M
 
Description Case-Based Reasoning produces a solution based on similarities to previous cases. New case solutions result from the placement of similarities with old cases. In this reseach the authors applied CBR to diagnose tuberculosis. System knowledge sources are obtained by collecting medical records of tuberculosis patients in 2014-2016. Calculation of similarity values using the K-Nearest Neighbor algorithm with a thereshold value of 80%. This system can diagnose 3 types of tuberculosis based on 25 symptoms. The system output consists of the type of tuberculosis based on the symptoms experienced by the patient, treatment solutions and presentation of similarities between new cases and old cases. Based on the results of testing with 51 cases the results: (a) testing with 3 new case scenarios obtained the accuracy of each system for data scenarios obtained by 31 training data (60% of 51 cases) and 20 test data (40% of 51 cases) accuracy is 63%, the second scenario accuracy obtained with 35 training data (70% of 51 cases) and 16 test data (30% of 51 cases) accuracy is 69.2% and the third scenario accuracy obtained with 41 training data (80% of 51 cases) and 10 test data (20% of 51 cases) accuracy is 90%. (b) The results of testing of the old cases in the case base obtained 100% accuracy of the system.
—Penalaran Berbasis Kasus menghasilkan solusi berdasarkan kemiripan terhadap kasus-kasus yang pernah terjadi sebelumnya. Solusi kasus baru dihasilkan dari pencocokan kemiripan dengan kasus lama. Pada penelitian ini penulis menerapkan CBR untuk mendignosa penyakit tuberkulosa. Sumber pengetahuan sistem diperoleh dengan mengumpulkan berkas rekam medis pasien tuberkulosis pada tahun 2014-2016. Perhitungan nilai kemiripan menggunakan algoritma K-Nearest Neighbor dengan nilai batas kewajaran 80%. Sistem ini dapat mendiagnosis 3 jenis penyakit tuberkulosis berdasarkan 25 gejala yang ada. Luaran sistem berupa jenis penyakit tuberkulosis berdasarkan gejala yang dialami pasien, solusi pengobatan dan presentasi kemiripan antara kasus baru dan kasus lama. Berdasarkan hasil pengujian dengan 51 kasus TB didapatkan hasil: (a) pengujian dengan 3 skenario pengujian kasus baru didapatkan keakuratan sistem masing-masing untuk skenario pertama akurasi yang diperoleh dengan 31 data latih (60% dari 51 kasus) dan 20 data uji (40% dari 51 kasus) akurasinya sebesar 63%, skenario kedua akurasi yang diperoleh dengan 35 data latih (70% dari 51 kasus) dan 16 data uji (30% dari 51 kasus) akurasinya sebesar 69.2% dan skenario ketiga akurasi yg diperoleh dengan 41 data latih (80% dari 51 kasus) dan 10 data uji (20% dari 51 kasus) akurasinya sebesar 90%, (b) hasil pengujian terhadap kasus lama dalam basis kasus didapatkan keakuratan sistemsebesar 100%.
 
Publisher Universitas Nusa Cendana
 
Date 2019-11-27
 
Type info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
Peer-reviewed Article
 
Format application/pdf
 
Identifier http://ejurnal.undana.ac.id/jicon/article/view/1656
10.35508/jicon.v7i2.1656
 
Source J-Icon : Jurnal Komputer dan Informatika; Vol 7 No 2 (2019): Oktober 2019; 124-128
2654-4091
2337-7631
10.35508/jicon.v7i2
 
Language eng
 
Relation http://ejurnal.undana.ac.id/jicon/article/view/1656/1288
 
Rights Copyright (c) 2019 Jurnal Komputer dan Informatika (JICON)
http://creativecommons.org/licenses/by-nd/4.0
 

Contact Us

The PKP Index is an initiative of the Public Knowledge Project.

For PKP Publishing Services please use the PKP|PS contact form.

For support with PKP software we encourage users to consult our wiki for documentation and search our support forums.

For any other correspondence feel free to contact us using the PKP contact form.

Find Us

Twitter

Copyright © 2015-2018 Simon Fraser University Library