Time series forecasting of new cases and new deaths rate for COVID-19 using deep learning methods

被引:69
作者
Ayoobi, Nooshin [1 ]
Sharifrazi, Danial [2 ]
Alizadehsani, Roohallah [3 ]
Shoeibi, Afshin [4 ,5 ]
Gorriz, Juan M. [6 ]
Moosaei, Hossein [7 ]
Khosravi, Abbas [3 ]
Nahavandi, Saeid [3 ]
Chofreh, Abdoulmohammad Gholamzadeh [8 ]
Goni, Feybi Ariani [9 ]
Klemes, Jiri Jaromir [8 ]
Mosavi, Amir [10 ,11 ]
机构
[1] Savitribai Phule Pune Univ, Dept Math, Pune 411007, Maharashtra, India
[2] Islamic Azad Univ, Sch Tech & Engn, Dept Comp Engn, Shiraz, Iran
[3] Deakin Univ, Inst Intelligent Syst Res & Innovat IISRI, Waurn Ponds, Vic 3217, Australia
[4] Ferdowsi Univ Mashhad, Comp Engn Dept, Mashhad, Razavi Khorasan, Iran
[5] KN Toosi Univ Technol, Biomed Data Acquisit Lab, Fac Elect & Comp Engn, Tehran, Iran
[6] Univ Granada, Dept Signal Theory Networking & Commun, Granada, Spain
[7] Univ Bojnord, Fac Sci, Dept Math, Bojnurd, Iran
[8] Brno Univ Technol VUT Brno, Fac Mech Engn, NETME Ctr, Sustainable Proc Integrat Lab SPIL, Tech 2896-2, Brno 61669, Czech Republic
[9] Brno Univ Technol VUT Brno, Fac Business & Management, Dept Management, Kolejni 2906-4, Brno 61200, Czech Republic
[10] Obuda Univ, John von Neumann Fac Informat, H-1034 Budapest, Hungary
[11] Norwegian Univ Life Sci, Sch Business & Econ, N-1430 As, Norway
关键词
Long Short Term Memory (LSTM); Convolutional Long Short Term Memory (Conv-LSTM); Gated Recurrent Unit (GRU); Bidirectional New Cases of COVID-19; New Deaths of COVID-19; COVID-19; Prediction; Deep learning; Machine learning; MODEL;
D O I
10.1016/j.rinp.2021.104495
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
The first known case of Coronavirus disease 2019 (COVID-19) was identified in December 2019. It has spread worldwide, leading to an ongoing pandemic, imposed restrictions and costs to many countries. Predicting the number of new cases and deaths during this period can be a useful step in predicting the costs and facilities required in the future. The purpose of this study is to predict new cases and deaths rate one, three and seven-day ahead during the next 100 days. The motivation for predicting every n days (instead of just every day) is the investigation of the possibility of computational cost reduction and still achieving reasonable performance. Such a scenario may be encountered in real-time forecasting of time series. Six different deep learning methods are examined on the data adopted from the WHO website. Three methods are LSTM, Convolutional LSTM, and GRU. The bidirectional extension is then considered for each method to forecast the rate of new cases and new deaths in Australia and Iran countries. This study is novel as it carries out a comprehensive evaluation of the aforementioned three deep learning methods and their bidirectional extensions to perform prediction on COVID-19 new cases and new death rate time series. To the best of our knowledge, this is the first time that Bi-GRU and Bi-Conv-LSTM models are used for prediction on COVID-19 new cases and new deaths time series. The evaluation of the methods is presented in the form of graphs and Friedman statistical test. The results show that the bidirectional models have lower errors than other models. A several error evaluation metrics are presented to compare all models, and finally, the superiority of bidirectional methods is determined. This research could be useful for organisations working against COVID-19 and determining their long-term plans.
引用
收藏
页数:15
相关论文
共 44 条
  • [1] Alizadehsani R., 2021, ARXIV PREPRINT ARXIV
  • [2] Risk factors prediction, clinical outcomes, and mortality in COVID-19 patients
    Alizadehsani, Roohallah
    Sani, Zahra Alizadeh
    Behjati, Mohaddeseh
    Roshanzamir, Zahra
    Hussain, Sadiq
    Abedini, Niloofar
    Hasanzadeh, Fereshteh
    Khosravi, Abbas
    Shoeibi, Afshin
    Roshanzamir, Mohamad
    Moradnejad, Pardis
    Nahavandi, Saeid
    Khozeimeh, Fahime
    Zare, Assef
    Panahiazar, Maryam
    Acharya, U. Rajendra
    Islam, Sheikh Mohammed Shariful
    [J]. JOURNAL OF MEDICAL VIROLOGY, 2021, 93 (04) : 2307 - 2320
  • [3] Arun SS, 2020, 2020 4 INT C INT COM 2020 4 INT C INT COM
  • [4] Asgharnezhad H, 2020, ARXIV PREPRINT ARXIV
  • [5] Babaei A, 2021, APPL COMPUT MATH-BAK, V20, P140
  • [6] Mathematical analysis of a stochastic model for spread of Coronavirus
    Babaei, A.
    Jafari, H.
    Banihashemi, S.
    Ahmadi, M.
    [J]. CHAOS SOLITONS & FRACTALS, 2021, 145
  • [7] A mathematical model to examine the effect of quarantine on the spread of coronavirus
    Babaei, A.
    Ahmadi, M.
    Jafari, H.
    Liya, A.
    [J]. CHAOS SOLITONS & FRACTALS, 2021, 142
  • [8] DC programming and DCA for parametric-margin ν-support vector machine
    Bazikar, Fatemeh
    Ketabchi, Saeed
    Moosaei, Hossein
    [J]. APPLIED INTELLIGENCE, 2020, 50 (06) : 1763 - 1774
  • [9] A fractional-order model describing the dynamics of the novel coronavirus (COVID-19) with nonsingular kernel
    Boudaoui, Ahmed
    Moussa, Yacine El Hadj
    Hammouch, Zakia
    Ullah, Saif
    [J]. CHAOS SOLITONS & FRACTALS, 2021, 146
  • [10] Active Monitoring of Persons Exposed to Patients with Confirmed COVID-19-United States, January-February 2020
    Burke, Rachel M.
    Midgley, Claire M.
    Dratch, Alissa
    Fenstersheib, Marty
    Haupt, Thomas E.
    Holshue, Michelle
    Ghinai, Isaac
    Jarashow, M. Claire
    Lo, Jennifer
    McPherson, Tristan D.
    Rudman, Sara
    Scott, Sarah
    Hall, Aron J.
    Fry, Alicia M.
    Rolfes, Melissa A.
    [J]. MMWR-MORBIDITY AND MORTALITY WEEKLY REPORT, 2020, 69 (09): : 245 - 246