Covid-19'un Yayılım Tahminine Yönelik Makine Öğrenmesi ve Derin Öğrenme Tabanlı Karşılaştırmalı Bir Analiz: Türkiye İçin Örnek Bir Çalışma.

UTKU, Anıl; CAN, Ümit

Epidemics have occurred in various time periods throughout history and have caused serious damage to human communities. Today, the modern version of these epidemics, Covid-19, has caused millions of people to die and have health problems as well. The entire world is making incredible efforts to combat the spread of this deadly disease in terms of infrastructure, finance, data resources, protective equipment, life-threatening treatments, and many other resources. Researchers are developing mathematical models to analyze this epidemic situation using data shared across the country. Countries are trying to fight this epidemic depending on the speed of the epidemic. In this study, an LSTM-based prediction model was created to predict the number of cases and deaths in Turkey. Six machine learning methods, including popular deep learning methods, RF, SVM, XGBoost, MLP, CNN, and RNN, were used to measure the prediction success of this model. The LSTM model has been the most successful model in predicting the number of cases, with MSE: 16670823,040 RMSE: 4082,991 MAE: 2543,651 R2: 0.975, and in predicting the number of deaths, MSE: 331,351 RMSE: 18,203 MAE: 14,891 R2: 0.740.

Firat University Journal of Engineering Science / Fırat Üniversitesi Mühendislik Bilimleri Dergisi, 2022, Vol 34, Issue 2, p709

1308-9072

Academic Journal

10.35234/fumbd.1125609