Please use this identifier to cite or link to this item:
|Title:||An approximate method for constructing linear regression models for prediction of a severe bronchial asthma course|
|Authors:||Pihnastyi, O. M.|
Kozhyna, O. S.
|Keywords:||regression model; regression model; child|
|Publisher:||Institute of Electrical and Electronics Engineers, Inc., USA|
|Citation:||Pihnastyi O. An approximate method for constructing linear regression models for prediction of a severe bronchial asthma course / O. Pihnastyi, O. Kozhyna // 2021 IEEE 12th International Conference on Electronics and Information Technologies (ELIT) : proc., 19-21 May 2021, Lviv, Ukraine. – P. 255-260.|
|Abstract:||Diagnosis of a bronchial asthma course in children determines further preventive activities and personalized approaches to treating a child with such pathology. An uncontrolled type of asthma course requires conducting a careful analysis of the factors affecting the formation of severe forms of the disease. The use of linear regression models is a widespread approach that helps to calculate the probability of severe bronchial asthma or the uncontrolled course of the disease development. During this study, 90 children aged from 6 to 18 years old were examined. Of these, there were 70 children suffering from bronchial asthma with different severity and 20 healthy children. The examination included an interviewing of patients as well as a definition of clinical features and results of clinical and laboratory examination in the disease course. 142 factors were analyzed to build a three-parameter model. A correlation ratio and numerical characteristics of model regressors were calculated. Conditions of the use of approximate linear regression models were shown and the model accuracy was estimated. A technique of an approximate model of linear regression building both in dimensional and nondimensional forms was considered. The relationship among model ratios was shown.|
|Appears in Collections:||Кафедра "Інтернет речей"|
Files in This Item:
|Pihnastyi_An_approximate_method_2021.pdf||166,84 kB||Adobe PDF||View/Open|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.