TIME SERIES ANALYSIS TO FORECAST COVID-19 CASES IN CENTRAL ASIA

No Thumbnail Available

Date

2021

Journal Title

Journal ISSN

Volume Title

Publisher

СДУ хабаршысы - 2021

Abstract

Abstract. According to the study, the cases are expected to increase in the upcoming days. An exponential rise in the number of cases is also noticeable in a time series analysis. The current forecast models are expected to help the government and medical professionals plan for future circumstances and enhance healthcare system readiness. The proposed study employs a support vector regression model for forecasting the overall number of deaths, recovered cases, cumulative number of reported cases, and regular case count. The starting information is retrieved from the Ist of March to the 30th of April, 2021 (61 Days). The model predicts deaths, recoveries, and the total number of confirmed cases with an accuracy of over 97 percent, and regular new cases with an accuracy of 87 percent. The findings point to a Gaussian reduction in the number of cases, which may take another 3 to 4 months to reach the bare minimum of no new cases registered.

Description

Keywords

COVID 19, Support vector regression, Data analysis, Central Asia, СДУ хабаршысы - 2021, №1

Citation

R. Tabarek, B. Murat, A. Kaiyr, N. Smadyarov, D. Issa / TIME SERIES ANALYSIS TO FORECAST COVID-19 CASES IN CENTRAL ASIA / СДУ хабаршысы - 2021