바로가기메뉴

본문 바로가기 주메뉴 바로가기

ACOMS+ 및 학술지 리포지터리 설명회

  • 한국과학기술정보연구원(KISTI) 서울분원 대회의실(별관 3층)
  • 2024년 07월 03일(수) 13:30
 

logo

A Study on the prediction dyspnea-induced attributes of linear regression-based Article

인공지능연구 / Korean Journal of Artificial Intelligence, (E)2508-7894
2018, v.6 no.2, pp.17-22
https://doi.org/https://doi.org/10.24225/kjai.2018.6.2.17
Lee, Kwang-Keun (Department of Social Welfare, Kyungdong University)
Jeon, Gyu-Hyeon (Department of Medical IT Marketing, Eulji University)

Abstract

According to the World Health Organization, the top 10 causes of death worldwide include heart disease. Heart diseases include coronary disease, which induces acute myocardial infarction. Ticagrelor drugs are being used to treat acute alliances, but it has become difficult to breathe due to the drugs. In a related study, Tobias predicted that uric acid causes acute respiratory distress independently of other factors, including BNP. And in the Ahmad study, serum uric acid numbers were related to the left ventricle depending on the level of uric acid. Experimental data are data used after 155 patients who received coronary intervention took ticagrelor. The research methods were leveraged by gradient decent algorithm and linear regression. In order to avoid overfitting in the experiment, training data and test data were separated into 70 and 30 percent respectively. The experimental results lacked the predictability of other attributes except DT in the correlation coefficient and crystal coefficient. However, all attributes related to dyspnea other than DT are determined to be related to causing relaxation of the heart in the left ventricle. Therefore, the attribute causing dyspnea is determined to be an attribute causing relaxation of the heart of the DT and left ventricle.

keywords
Ticagrelor, Dyspnea Myocardial Infarction, Linear Regression, Uric Acid

인공지능연구