TY  - JOUR
TI  - A Study on the Estimation of COVID-19 Daily Cases and Reproduction Number Using Adaptive Kalman Filter for USA, Brazil, Germany, India, Russia, Italy, Spain, United Kingdom, France, Turkey
AB  - Objective: In the literature, non-linear mathematical growth models are often used to estimate the number of coronavirus disease-2019 (COVID-19) cases. Specific algorithms such as mathematical optimization technique need to be employed for parameter estimation. In this work, a novel method to estimate COVID-19 daily cases and reproduction number is proposed for COVID-19. Material and Methods: In this study, the daily number of COVID- 19 cases between January 01 and November 16, 2020 has been estimated online via AR(1) (autoregressive time-series model of order 1) and the adaptive Kalman filter (AKF). After calculating the estimate for daily cases, the reproduction number estimate was obtained. Results: It is quite a simple method to model the daily case number by time series with the time-varying parameter AR(1) stochastic process and estimated the time-varying parameter with online AKF. The method is online. Only the data points on the last day are sufficient. Conclusion: The COVID-19 data have been modeled in state space, and the AKF has been employed to estimate the number of daily cases. The estimation results were obtained for the number of daily cases using the AR(1) model. Since the estimation using the AR(1) stochastic process does not require any other modeling assumption, it is a simple approach to model the daily case number time series with the time-varying parameter AR(1) stochastic process and estimated the time-varying parameter with online AKF. We suggest that the simplest method for the reproduction number estimation will be obtained by modeling the daily case via an AR(1) model.
AU  - Demirtas, Hakan
AU  - ÖZBEK, Levent
DO  - 10.5336/biostatic.2020-80186
PY  - 2021
JO  - Türkiye Klinikleri Biyoistatistik Dergisi
VL  - 13
IS  - 1
SN  - 1308-7894
SP  - 91
EP  - 102
DB  - TRDizin
UR  - http://search/yayin/detay/491691
ER  -