Молимо вас користите овај идентификатор за цитирање или овај линк до ове ставке: https://scidar.kg.ac.rs/handle/123456789/15839
Пун извештај метаподатака
Поље DC-а ВредностЈезик
dc.contributor.authorSustersic, Tijana-
dc.contributor.authorBlagojevic, Andjela-
dc.date.accessioned2023-02-08T15:54:55Z-
dc.date.available2023-02-08T15:54:55Z-
dc.date.issued2022-
dc.identifier.issn--
dc.identifier.urihttps://scidar.kg.ac.rs/handle/123456789/15839-
dc.description.abstractAlthough ML has been examined for a variety of epidemiological and clinical concerns, as well as for COVID-19 survival prediction, there is a notable lack of research dealing with ML utilization in predicting disease severity changes during the course of the disease. This chapter encompasses two approaches in predicting COVID-19 spread—personalized model for predicting disease development in infected individual patients and an epidemiological model for predicting disease spread in population. Personalized model uses XGboost for the classification of infected individuals into four different groups based on the values of blood biomarkers analyzed by Gradient boosting regressor and chosen as biomarkers with the highest effect on the classification of COVID-19 patients. The epidemiological model includes two proposed methods—differential equation-based SEIRD model and an LSTM deep learning model. Proposed models can be used as tools useful in the research and control of infectious illnesses and in reducing the burden on the health system.-
dc.rightsinfo:eu-repo/semantics/restrictedAccess-
dc.sourceCardiovascular and Respiratory Bioengineering-
dc.titleArtificial intelligence approach toward analysis of COVID-19 development—Personalized and epidemiological model-
dc.typebookPart-
dc.identifier.doi10.1016/B978-0-12-823956-8.00013-4-
dc.identifier.scopus2-s2.0-85138351621-
Налази се у колекцијама:Faculty of Engineering, Kragujevac

Број прегледа

675

Број преузимања

8

Датотеке у овој ставци:
Датотека Опис ВеличинаФормат 
PaperMissing.pdf
  Ограничен приступ
29.86 kBAdobe PDFСличица
Погледајте


Ставке на SCIDAR-у су заштићене ауторским правима, са свим правима задржаним, осим ако није другачије назначено.