@article{scholars15655,
       publisher = {Scibulcom Ltd.},
          volume = {22},
            note = {cited By 2},
            year = {2021},
          number = {4},
         journal = {Journal of Environmental Protection and Ecology},
           pages = {1662--1675},
           title = {Performance comparison of different machine learning algorithms on a time-series of covid-19 data: A case study for Saudi Arabia},
             url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85115932423&partnerID=40&md5=ada0b8889b5821a0d4bd0fcb6dc681e4},
        abstract = {In this study we have applied several machine learning algorithms to analyse time-series data related to COVID-19 in Saudi Arabia. We retrieved the data from the official health website of Saudi Arabia for the period March 2nd 2020, to November 27st 2020. Several machine learning models and related algorithms were developed for prediction of total cases and total deaths. The COVID-19 data have been considered as a time-series dataset and the prediction capability of three machine learning methodologies, linear regression, support vector regression and Gaussian process regression, have been compared. When comparing all models based on R2 and RMSE values, it can be inferred that the linear regression and Gaussian process regression models were the most robust models for the prediction of total cases, and total deaths while SVM models were shown less prediction capabilities. Prediction of total cases and total deaths are obtained by taking previous 14 days of time series data as the input to the machine learning algorithms developed in this paper. This study can be helpful in analysing the capabilities of machine learning methodologies for time-series data-sets as well as helping governments in the decision making process for mitigation of the pandemic. {\^A}{\copyright} 2021, Scibulcom Ltd.. All rights reserved.},
        keywords = {algorithm; COVID-19; decision making; Gaussian method; infectious disease; machine learning; regression analysis; time series, Saudi Arabia},
          author = {Ahmad, M. T. and Qaiyum, S. and Alamri, A. and Islam, S.},
            issn = {13115065}
}