Bayesian Networks Modeling Using Partial Least Squares Approach to Predict Stroke Disease

Tarumanagara International Conference on the Applications of Technology and Engineering

工业技术（总论）

Trisnawarman, Dedi^1 ; Sutrisno, Tris^2 ; Perdana, Novario Jaya^1 ; Sugesti, S.^3

Information Systems Department, Faculty of Information Technology Universitas Tarumanagara, Indonesia^1

Informatics Engineering Department, Faculty of Information Technology Universitas Tarumanagara, Indonesia^2

STMIK Dharma Putra Tangerang, Indonesia^3

关键词: Bayesian Networks (bns);    Causes of death;    Decision making models;    Hospital medical records;    Partial least square (PLS);    Pls approaches;    Stroke;    Young peoples;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/508/1/012121/pdf
DOI  :  10.1088/1757-899X/508/1/012121

学科分类：工业工程学

来源: IOP

【 摘 要 】

Stroke is one of the leading causes of death in Indonesia. However, the incidence of stroke continues to increase. It could affect not only old people but also young people. In Indonesia, it is estimated that 500,000 people suffer from stroke every year, and about 25% or 125,000 people die and the rest suffer minor or severe defects. This study aims to build a decision-making model to help diagnose the possibility of stroke attacks. The model combines Bayesian Networks (BNs) and Partial Least Square (PLS) methods for predicting the attack on suspected patients. The model have been tested using PLS-PM approach. Hospital medical record was used as the testing data with the help of expert verification. The results showed that the model structure of BNs built on expert assumptions can be tested using the PLS approach, and stroke disease can be predicted using interrelated indicators in the model structure of BNs.