PREPROCESSING APPROACH FOR TUBERCULOSIS DNA CLASSIFICATION USING SUPPORT VECTOR MACHINES (SVM)

Anshori, Mochammad and Mahmudy, Wayan Firdaus and Supianto, Ahmad Afif (2019) PREPROCESSING APPROACH FOR TUBERCULOSIS DNA CLASSIFICATION USING SUPPORT VECTOR MACHINES (SVM). JITeCs (Journal of Information Technology and Computer Science), 4 (3). pp. 233-240. ISSN 2540-9433; E-ISSN 2540-9824

[img]
Preview
Text
Preprocessing Approach for Tuberculosis DNA.pdf

Download (964kB) | Preview
[img]
Preview
Text
Peer Review.pdf

Download (522kB) | Preview
[img]
Preview
Text
Uji Plagiasi.pdf

Download (1MB) | Preview

Abstract

Tuberculosis is a disease that caused by the mycobacterium tuberculosis virus. Tuberculosis is very dangerous and it is one of the top 10 causes of the death in the world. In its detection, errors often occur because it is similar to the other lung disease. The challenge is how to get the best detection system for classification of Tuberculosis using Deoxyribo Nucleic Acid (DNA) sequence data from mycobacterium tuberculosis. One way to do the detection is using machine learning algorithm which is Support Vector Machines (SVM). Before making a detection for Tuberculosis DNA Classification, it is necessarry to preprocess the DNA datasets first. Preprocessing method that we used in this research are using k-Mer for feature extraction, then processed with TF-IDF to transform it into numerical value and uniform the data length. Not Only that, because the DNA datasets is very large so dimension reduction is really needed and we used Linear Discriminant Analysis (LDA). Classification of Tubercolosis DNA will be done using Support Vector Machine (SVM) method with the best preprocesing method. So, in this research to get the best detection for DNA classification will be tested several experiment of parameters value from the method that we used in this research. The overall result based on the experiment of this research, the best k of k-Mer value is 5 that produce accuracy, precision, recall, F Score, and MCC are 0.927, 0.927, 0.920, 0.875.

Item Type: Article
Contributors:
ContributionNameNIDN / NIDKEmail
ReviewerMahmudy, Wayan FirdausNIDN0019097205UNSPECIFIED
ReviewerTolle, HermanNIDN0023087401UNSPECIFIED
Divisions: Informatics Study Program
Depositing User: Yacobus Sudaryono
Date Deposited: 08 Jul 2022 08:55
Last Modified: 04 Sep 2023 03:05
URI: http://repository.itsk-soepraoen.ac.id/id/eprint/710

Actions (login required)

View Item View Item