Makine Öğrenmesi Yöntemleri ile Türkiye’de Covid-19’a İlişkin Günlük Vaka, Ağır Hasta, Vefat ve İyileşen Sayısı Tahmini

Covid-19 içinde bulunduğumuz yüzyılın ilk pandemisidir ve bundan önceki pandemilere kıyasla süresi, neden olduğu can kaybı, yarattığı psikolojik, sosyolojik ve ekonomik etkileri dolayısıyla farklılık göstermektedir. Bu süreçte virüs pek çok varyant üretmiştir ve üretmeye de devam etmektedir. Dünya üzerindeki hareketliliğin sıklığı ve miktarı düşünüldüğünde, bu durumun yakın gelecekte değişmesi mümkün gözükmemektedir. Pandeminin gidişatını anlamak, bundan sonraki olası pandemiler için hazırlıklı olmak konusunda faydalı olacaktır. Bu amaçla, T.C. Sağlık Bakanlığı tarafından yayınlanan günlük veri incelenmiş, farklı veri grupları üzerinde gerek özelliklerini anlama gerekse geleceğe yönelik tahmin gerçekleştirme amacıyla, güncel bir yaklaşım olan makine öğrenmesi yöntemleri kullanılmıştır. Kul-lanılan veri grupları oldukça karmaşık birer zaman serisi yapısındadır ve günlük vaka sayısı, ağır hasta sayısı, günlük vefat sayısı ve günlük iyileşen sayısı olarak seçilmiştir. Polinom regresyon, en küçük kareler polinom uyumu ve kübik eğri uyumu sonuçları ile tahminler bu makalede incelenmiştir. Sonuçlar gerek grafikler yoluyla gerekse zaman serisi tahmininde kabul görmüş bir performans kriteri olan Canberra uzaklığının ortalama, medyan, standart sapma ve top-lam değerleriyle, sayısal olarak belirtilmiştir. Yukarıda belirtilen dört zaman serisi için en iyi sonuçların, kübik eğri uyumu yöntemiyle alındığı görülmektedir. Tahminlerde kullanılan eğrilerin dereceleri, zaman serisine göre değişiklik göstermektedir. Elde edilen tahmin sonuçları, zaman serisine bağlı olarak değişen yüksek doğruluk oranı sağlamıştır.

Anahtar Kelimeler:

Covid-19, makine öğrenmesi, en küçük kareler polinom uyumu, kübik eğri uyumu, polinom regresyon

Estimation of Daily Cases, Deaths, Serious Patients and Recovering Patients of Covid-19 in Turkey with Machine Learning Methods

Covid-19 is the first pandemic of the century and differs from previous pandemics due to its duration, loss of life, and psychological, sociological and economic effects. In this process, the virus has produced and continues to produce many variants. Considering the frequency and amount of mobility on Earth, this situation does not seem likely to change soon. Understanding the course of the pandemic will be helpful in being prepared for the next possible pandemics. To this end, the daily data published by the Turkish Ministry of Health was examined, and machine learn-ing methods, which is an up-to-date approach, were used to understand the features and make predictions for the future on different data groups. The data groups used are in a very complex time series structure and were chosen as the number of daily cases, severe patients, deaths, and recoveries per day. The results of polynomial regression, least squares polynomial fit, and cubic spline fit, and estimations are shown in this article. The results are presented graph-ically, and by means of an accepted performance criterion in time series estimation, namely by the mean, median, standard deviation, and total values of the Canberra distance. It is seen that the best results for the time series mentioned above are obtained by the cubic spline fit method. The degrees of the curves used in the estimations vary according to the time series. The estimation results obtained provided a high accuracy rate that varies depending on the time series.

Keywords:

Covid 19, machine learning, cubic spline fit, least squares polynomial fit, polynomial regression,

PDF

___

Aggarwal, C. C. (2018). Neural Networks and Deep Learning. Springer.
ArunKumar, K. E., Kalaga, D. V., Sai Kumar, C. M., Chilkoor, G., Kawaji, M., ve Brenza, T. M. (2021). Forecasting the dynamics of cumulative COVID-19 cases (confirmed, recovered and deaths) for top-16 countries using statistical machine learning models: Auto-Regressive Integrated Moving Average (ARIMA) and Seasonal Auto-Regressive Integrated Moving Averag. Applied Soft Computing, 103(December 2019), 107161. https://doi.org/10.1016/j.asoc.2021.107161
Bhadana, V., Jalal, A. S., ve Pathak, P. (2020). A Comparative Study of Machine Learning Models for Covid-19 Prediction in India. IEEE 4th Conference on Information ve Communication Technology (CICT). https://doi.org/10.1109/CICT51604.2020.9312112
Dos Santos Gomes, D. C., ve De Oliveira Serra, G. L. (2021). Machine Learning Model for Computational Tracking and Forecasting the COVID-19 Dynamic Propagation. IEEE Journal of Biomedical and Health Informatics, 25(3), 615–622. https://doi.org/10.1109/JBHI.2021.3052134
Ertel, W. (2017). Introduction to Artificial Intelligence (2. baskı). Springer.
Gambhir, E., Jain, R., Gupta, A., ve Tomer, U. (2020). Regression Analysis of COVID-19 using Machine Learning Algorithms. 2020 International Conference on Smart Electronics and Communication, Icosec, 65–71. https://doi.org/10.1109/ICOSEC49089.2020.9215356
Gupta, V. K., Gupta, A., Kumar, D., ve Sardana, A. (2021). Prediction of COVID-19 confirmed, death, and cured cases in India using random forest model. Big Data Mining and Analytics, 4(2), 116–123. https://doi.org/10.26599/BDMA.2020.9020016
Harrell Jr., F. E. (2015). Regression Modeling Strategies (2. baskı). Springer.
Hazarika, B. B., ve Gupta, D. (2020). Modelling and forecasting of COVID-19 spread using wavelet-coupled random vector functional link networks. Applied Soft Computing Journal, 96, 106626. https://doi.org/10.1016/j.asoc.2020.106626 Kumar, N., ve Susan, S. (2021). Particle swarm optimization of partitions and fuzzy order for fuzzy time series forecasting of COVID-19. Applied Soft Computing, 110, 107611. https://doi.org/10.1016/j.asoc.2021.107611
Kumari, R., Kumar, S., Poonia, R. C., Singh, V., Raja, L., Bhatnagar, V., ve Agarwal, P. (2021). Analysis and predictions of spread, recovery, and death caused by COVID-19 in India. Big Data Mining and Analytics, 4(2), 65–75. https://doi.org/10.26599/BDMA.2020.9020013
Kurniawan, R., Abdullah, S. N. H. S., Lestari, F., Nazri, M. Z. A., Mujahidin, A., ve Adnan, N. (2020). Clustering and Correlation Methods for Predicting Coronavirus COVID-19 Risk Analysis in Pandemic Countries. 2020 8th International Conference on Cyber and IT Service Management, CITSM 2020. https://doi.org/10.1109/CITSM50537.2020.9268920
Leon, M. I., Iqbal, M. I., Azim, S. M., ve Al Mamun, K. A. (2021). Predicting COVID-19 infections and deaths in Bangladesh using Machine Learning Algorithms. 2021 International Conference on Information and Communication Technology for Sustainable Development, ICICT4SD 2021 - Proceedings, 70–75. https://doi.org/10.1109/ICICT4SD50815.2021.9396820
Mandayam, A. U., Rakshith, A. C., Siddesha, S., ve Niranjan, S. K. (2020). Prediction of Covid-19 pandemic based on Regression. Proceedings - 2020 5th International Conference on Research in Computational Intelligence and Communication Networks, ICRCICN 2020, 1–5. https://doi.org/10.1109/ICRCICN50933.2020.9296175
Ramchandani, A., Fan, C., ve Mostafavi, A. (2020). DeepCOVIDNet: An Interpretable Deep Learning Model for Predictive Surveillance of COVID-19 Using Heterogeneous Features and Their Interactions. IEEE Access, 8, 159915–159930. https://doi.org/10.1109/ACCESS.2020.3019989
Rustam, F., Reshi, A. A., Mehmood, A., Ullah, S., On, B. W., Aslam, W., ve Choi, G. S. (2020). COVID-19 Future Forecasting Using Supervised Machine Learning Models. IEEE Access, 8, 101489–101499. https://doi.org/10.1109/ACCESS.2020.2997311
Sevli, O., ve Başer, V. G. (2020). Covid- 19 Salgınına Yönelik Zaman Serisi Verileri ile Prophet Model Kullanarak Makine Öğrenmesi Temelli Vaka Tahminlemesi Machine Learning Based Case Estimation Using Prophet Model with Time Series Data for Covid-19 Outbreak. 19, 827–835. https://doi.org/10.31590/ejosat.766623
Singh, M., ve Dalmia, S. (2020). Prediction of number of fatalities due to Covid-19 using Machine Learning. 2020 IEEE 17th India Council International Conference, INDICON 2020. https://doi.org/10.1109/INDICON49873.2020.9342390
T.C. Sağlık Bakanlığı. (2021). Covid 19. https://covid19.saglik.gov.tr/
Theodoridis, S., ve Koutroumbas, K. (2009). Pattern Recognition (4th Editio). Academic Press, Elsevier.
Vakula Rani, J., ve Jakka, A. (2020). Forecasting COVID-19 cases in India using machine learning models. Proceedings of the International Conference on Smart Technologies in Computing, Electrical and Electronics, ICSTCEE 2020, 466–471. https://doi.org/10.1109/ICSTCEE49637.2020.9276852
World Health Organization. (2021). Data Table. https://covid19.who.int/
Yang, Z., ve Chen, K. (2020). Machine Learning Methods on COVID-19 Situation Prediction. Proceedings - 2020 International Conference on Artificial Intelligence and Computer Engineering, ICAICE 2020, 78–83. https://doi.org/10.1109/ICAICE51518.2020.00021
Yudistira, N., Sumitro, S. B., Nahas, A., ve Riama, N. F. (2021). Learning where to look for COVID-19 growth: Multivariate analysis of COVID-19 cases over time using explainable convolution–LSTM. Applied Soft Computing, 109, 107469. https://doi.org/10.1016/j.asoc.2021.107469