استخدام نموذج الانحدار اللوجستي لتحديد أعراض الإصابة بفيروس كورونا

Abstract EN

The main goal of the research is to identify the factors affecting infection with the COVID-19, and to form a binary logistic regression model to predict the probability of the person being infected or not with the virus.

this was done through the binary logistic regression model through a qualitative dependent variable (infected or not infected), and a set of independent variables such as age, temperature, diarrhea, headache, shortness of breath, sore throat, chronic diseases, mixtures or non-mixtures, sense of smell, taste and cough.

the researcher found the significance of the effect of temperature, age, diarrhea, headache, smell and taste, contact with a sick or non-contact person, and non significance of both sore throats, shortness of breath, chronic diseases, and cough are shown.

the order of the variables according to the relative importance and their impact on infection with the virus was the first place, the sense of smell and taste, at 47.6 %, followed by diarrhea 14.62 %, followed by age 9.28 %, followed by temperature 9.124 %, followed by headache % 4.977 and finally contact or non-contact with a rate of 3.92 %.

the independent variables contribute by 79.4 % in explaining the changes that occur in the dependent variable and contribute by 59 % to improving the quality of the prediction of the model and contribute 89.6 % to the correct classification of people.