Data Analytics for Predicting COVID-19 Cases in Top Affected Countries: Observations and Recommendations
Abstract
:1. Introduction
2. Literature Review
2.1. Bibliographic Summary for COVID-19
2.2. COVID-19 Related Works
2.3. Data Analytics
3. Research Gaps and Contribution
4. Data Analytics for Predicting New Daily Cases of COVID-19
- Step 1:
- Collecting the data. The data have been collected from online websites, including “Worldometers” [39], “Our World in Data” [40], “World Bank Open Data” [41], and the official website of the World Health Organization (WHO). Besides, human development reports have been used to pick other kind of information, like median age and education index [42]. The scope of this study includes collecting data for about 189 countries/territories by focusing on two types of data: main data and other external factors. The main data include considering the number of confirmed coronavirus disease cases/day, the number of deaths due to coronavirus disease/day, and the total number of confirmed cases [39,40]. The external factors, on the other hand, include considering the factors that affect the spread of coronavirus disease. Note that the data have been collected for about 224 days, from 31 December 2019 until 10 August 2020. This leads naturally to set the size of data at 224.
- Step 2:
- Preprocessing the data. While collecting the data, it was observed that data were not always available for the whole 189 countries/territories. To alleviate this situation, a refinement was performed by ruling out any countries/territories that suffer from data unavailability. This results in cancelling around 39 countries/territories, so that only 150 countries/territories have been considered. Our preliminary goal is to predict the new cases for all 150 countries/territories. However, this is not reasonable for two reasons. Firstly, it is not possible to present all the results in a single study due to page limitation. Secondly, it is computationally expensive to run this algorithm for 150 countries/territories. For the above reasons, we have limited our scope to considering the most affected countries in each continent. By doing so, the top five affected countries/territories have been considered from each continent. More details are presented in Section 5.
- Step 3:
- Identifying the input sets. These sets contain historical data of some information alongside the external factors. These sets can be outlined as follows:
- Main set, which includes two main information: the number of deaths due to coronavirus disease/day and the total number of confirmed cases;
- External factor set that comprises the factors that affect the spread of coronavirus including population [42], median age index [41], public and private healthcare expenditure [41], air quality as a CO2 trend [42], number of arrivals in the countries/territories [41], and education index [42]. There is another factor that should be considered, called seasonality. Before incorporating this factor in the model, it should be clarified here that, in most countries, we can find cities with different seasons. For example, Iran has four seasons in its different cities [43]. Other examples include USA, China, Saudi Arabia, and Egypt. This observation indicates that, to consider the seasonality using seasons as a factor, the cities should be the scope of the study. Since the scope of this study is not cities but the countries, seasonality factor using seasons themselves cannot be considered in our algorithm. To find a compromise for this situation, the month of collecting data is selected to capture the seasonality in the proposed algorithm. It should be noted that the main data for the daily COVID-19 cases have been collected from the website “Our World in Data”, corrected through the website “Worldometers”. Next, the data have been doublechecked and refined by the data from the official website of WHO. In addition, because the considerable predictors are diverted, and they are not available on one database, their data have been collected from several websites. In further details, the data that have been collected from the website “World Bank Open Data” are the median age, number of arrivals, and health expenditure as a percentage of GDP [41], while the education index has been collected from United Nations Development Programme [42].
- Step 4:
- Test of hypothesis using regression analysis. Since our study goal is to accurately predict the number of COVID-19 cases, we should focus on the most influential external factors. To do so, test of hypotheses using regression analysis should be conducted for each external factor. These hypotheses can be outlined as follows:
- Step 5:
- Designing the neural network structure. We have utilized the feedforward time-delay neural network as this structure has been commonly used in the literature due to its efficiency [44]. This network is composed of three main layers: input, hidden, and output. Regarding the activation function, the sigmoid function has been selected because it is efficient in reflecting the non-linear relationship among multiple factors.
- Step 6:
- Training the neural network. To achieve this goal, the supervised learning method has been adopted. In this method, 70% of the data have been used for training purposes, whereas the rest have been reserved for validation and testing purposes.
- Step 7:
- Predicting new cases of COVID-19. The trained data, known as the output of the network, have been used to predict the new cases of COVID-19 in the period from August 2020 until September 2020.
5. Experiments and Results
5.1. Test of Hypothesis Using Regression Analysis
5.2. Parameter Settings of NARX Neural Network-Based Algorithm
5.3. Performance of the NARX Neural Network-Based Algorithm
5.4. Performance Analysis
5.5. What Next in the Future?
- In most European countries, like Italy, UK, and Russia, the number of cases has already reached its peak before our future prediction. Based on this observation, we predict that the number of future COIVD-19 cases will decrease gradually during the period from August 2020 until September 2020. It is worth to mention that our predicted reduction in the number of COVID-19 cases appear during August 2020. The abovementioned reduction is because of strict compliance with the precaution guidelines established by WHO.
- In contrast to most of European countries, the situation in Spain and France is quite similar, as the number of cases has raised recently and formed another peak. In Spain, the second peak has been already formed, therefore, our algorithm predicts a gradual decrease during the period from August 2020 until September 2020. In France, the algorithm predicts a slight increase followed by a gradual decrease in the number of cases during the same period.
- In the case of USA, the number of COVID-19 cases has already formed its second peak by mid of July 2020. Therefore, we have predicted a slow reduction in the number of future COVID-19 cases. This prediction, in terms of reduction, appears in the USA during the first half of August 2020. It should be noted that this slow reduction is because of the dysfunction experienced in the healthcare system of the USA [59].
- In the case of Brazil, we observe that the peak has been reached. Then, we predict that the number of future COVID-19 cases will experience a wavy reduction during the period from August 2020 until September 2020. It is important to mention that the predicted slow reduction in future COVID-19 cases agrees with the actual reduction realized during August 2020. This wavy reduction is due to overlooking the social distancing instructions by most of Brazilian residents [60].
- In case of Canada, the number of COVID-19 cases has reached its peak since beginning of May 2020. Therefore, it is reasonable to predict a gradual decrease in the number of cases during the period from August 2020 until September 2020.
- In the case of Peru, we observe an increase in the number of COVID-19 cases by July 2020. Then, we predict that, during the period from August 2020 until September 2020, this increase will continue a bit before a decline in the number of future COVID-19 cases. This increase is due to the bad behavior of the people, so that the situation becomes even worse during those days [61].
- In the case of Ecuador, we predict the number of future COVID-19 cases will keep its wavy motion, meaning that the number will tend to zero and increase again. This wave appears because there is no transparency in the reported number of COVID-19 cases, meaning that the government has not disclosed the real number of COVID-19 cases [62].
- China is one of the few cases that has fully controlled the situation. This is apparent as the number of future COVID-19 cases is almost zero. Thanks to the Chinese government and medical system, strict typical quarantining measures have been implemented, thus, leading finally to overcoming this hard time.
- The situation in Turkey, Iran, and Saudi Arabia is like other European countries that have reached the peak. Our algorithm predicts a gradual reduction in the number of future COVID-19 cases, which has been realized during August 2020.
- The situation in India is completely different compared to the rest of Asian countries. This is because India is yet to reach its peak. Based on this observation, we predict that the increase in the number of COVID-19 cases will continue during the period from August 2020 until September 2020.
- Most of African countries are quite similar, except Morocco, as the peak of COVID-19 cases has already appeared. It is predicted that the future number of COVID-19 cases will decrease gradually, during August and September 2020. So far, the trend of our predicted graph has been realized in the aforementioned African countries.
- In contrast to the above-mentioned African counties, in Morocco, the peak of COVID-19 cases has not yet appeared. Therefore, the future number of COVID-19 cases will continue its increase during August and September 2020. It is important to mention that this increase has been realized during the first half of August 2020.
- It is recommended for the people living in the USA, Brazil, Ecuador, Peru, and India to strictly follow the precautions instruction recommended by WHO. This includes quarantining infected people, whereas healthy people should stay home to avoid COVID-19 infection, and when they go out, they should follow the rules of social distancing.
- It is recommended for the government and healthcare system of countries like the USA and Brazil to raise their private and public health expenditures to control the number of future COIVD-19 cases. In addition, penalties may be applied to the people who violate the instructions recommended by WHO.
- It is recommended for the government of Ecuador to release the correct number of COVID-19 cases so that the people can understand the severity of the situation and obey the health guidelines released by WHO.
- In the countries that fully or partially control the COVID-19 like China, it is recommended for the people to keep following the medical instructions. Otherwise, COVID-19 may come back in a mutated form causing another global pandemic.
6. Conclusions
Author Contributions
Funding
Acknowledgments
Conflicts of Interest
References
- Wang, D.; Hu, B.; Hu, C.; Zhu, F.; Liu, X.; Zhang, J.; Wang, B.; Xiang, H.; Cheng, Z.; Xiong, Y.; et al. Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus–Infected Pneumonia in Wuhan, China. JAMA 2020, 323, 1061–1069. [Google Scholar] [CrossRef] [PubMed]
- Shen, K.; Yang, Y.; Wang, T.; Zhao, D.; Jiang, Y.; Jin, R.; Zheng, Y.; Xu, B.; Xie, Z.; Lin, L.; et al. Diagnosis, treatment, and prevention of 2019 novel coronavirus infection in children: Experts’ consensus statement. World J. Pediatrics 2020, 16, 223–231. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Lauer, S.A.; Grantz, K.H.; Bi, Q.; Jones, F.K.; Zheng, Q.; Meredith, H.R.; Azman, A.S.; Reich, N.G.; Lessler, J. The Incubation Period of Coronavirus Disease 2019 (COVID-19) From Publicly Reported Confirmed Cases: Estimation and Application. Ann. Intern. Med. 2020, 172, 577–582. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Backer, J.A.; Klinkenberg, D.; Wallinga, J. Incubation period of 2019 novel coronavirus (2019-nCoV) infections among travellers from Wuhan, China, 20–28 January 2020. Eurosurveillance 2020, 25, 2000062. [Google Scholar] [CrossRef] [Green Version]
- Remuzzi, A.; Remuzzi, G. COVID-19 and Italy: What next? Lancet 2020, 395, 1225–1228. [Google Scholar] [CrossRef]
- Tuite, A.R.; Ng, V.; Rees, E.; Fisman, D. Estimation of COVID-19 outbreak size in Italy. Lancet Infect. Dis. 2020, 20, 537. [Google Scholar] [CrossRef] [Green Version]
- Tuite, A.R.; Bogoch, I.I.; Sherbo, R.; Watts, A.; Fisman, D.; Khan, K. Estimation of coronavirus disease 2019 (COVID-19) burden and potential for international dissemination of infection from Iran. Ann. Intern. Med. 2020, 172, 699–701. [Google Scholar] [CrossRef] [Green Version]
- Zhuang, Z.; Zhao, S.; Lin, Q.; Cao, P.; Lou, Y.; Yang, L.; He, D. Preliminary estimation of the novel coronavirus disease (COVID-19) cases in Iran: A modelling analysis based on overseas cases and air travel data. Int. J. Infect. Dis. 2020, 94, 29–31. [Google Scholar] [CrossRef]
- Ceylan, Z. Estimation of COVID-19 prevalence in Italy, Spain, and France. Sci. Total Environ. 2020, 729, 138817. [Google Scholar] [CrossRef]
- Batista, M. Estimation of the final size of the COVID-19 epidemic. medRxiv 2020. [Google Scholar] [CrossRef] [Green Version]
- Alboaneen, D.; Pranggono, B.; Alshammari, D.; Alqahtani, N.; Alyaffer, R. Predicting the Epidemiological Outbreak of the Coronavirus Disease 2019 (COVID-19) in Saudi Arabia. Int. J. Environ. Res. Public Health 2020, 17, 4568. [Google Scholar] [CrossRef] [PubMed]
- Papastefanopoulos, V.; Linardatos, P.; Kotsiantis, S. COVID-19: A Comparison of Time Series Methods to Forecast Percentage of Active Cases per Population. Appl. Sci. 2020, 10, 3880. [Google Scholar] [CrossRef]
- Hernandez-Matamoros, A.; Fujita, H.; Hayashi, T.; Perez-Meana, H. Forecasting of COVID19 per regions using ARIMA models and polynomial functions. Appl. Soft Comput. 2020, 96, 106610. [Google Scholar] [CrossRef] [PubMed]
- Qin, L.; Sun, Q.; Wang, Y.; Wu, K.F.; Chen, M.; Shia, B.C.; Wu, S.Y. Prediction of Number of Cases of 2019 Novel Coronavirus (COVID-19) Using Social Media Search Index. Int. J. Environ. Res. Public Health 2020, 17, 2365. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Li, C.; Chen, L.J.; Chen, X.; Zhang, M.; Pang, C.P.; Chen, H. Retrospective analysis of the possibility of predicting the COVID-19 outbreak from Internet searches and social media data, China, 2020. Eurosurveillance 2020, 25, 2000199. [Google Scholar] [CrossRef]
- Li, J.; Xu, Q.; Cuomo, R.; Mackey, T.; Purushothaman, V. Data Mining and Content Analysis of the Chinese Social Media Platform Weibo During the Early COVID-19 Outbreak: Retrospective Observational Infoveillance Study. JMIR Public Health Surveill. 2020, 6, e18700. [Google Scholar] [CrossRef] [Green Version]
- Chen, F.-M.; Feng, M.-C.; Chen, T.-C.; Hsieh, M.-H.; Kuo, S.-H.; Chang, H.-L.; Yang, C.-J.; Chen, Y.-H. Big data integration and analytics to prevent a potential hospital outbreak of COVID-19 in Taiwan. J. Microbiol. Immunol. Infect. 2020. [Google Scholar] [CrossRef]
- Zhou, C.; Su, F.; Pei, T.; Zhang, A.; Du, Y.; Luo, B.; Cao, Z.; Wang, J.; Yuan, W.; Zhu, Y. COVID-19: Challenges to GIS with big data. Geogr. Sustain. 2020, 1, 77–87. [Google Scholar] [CrossRef]
- Wieczorek, M.; Siłka, J.; Woźniak, M. Neural network powered COVID-19 spread forecasting model. Chaos Solitons Fractals 2020, 140, 110203. [Google Scholar] [CrossRef]
- Pinter, G.; Felde, I.; Mosavi, A.; Ghamisi, P.; Gloaguen, R. COVID-19 Pandemic Prediction for Hungary; A Hybrid Machine Learning Approach. Mathematics 2020, 8, 890. [Google Scholar] [CrossRef]
- Bragazzi, N.L.; Dai, H.; Damiani, G.; Behzadifar, M.; Martini, M.; Wu, J. How Big Data and Artificial Intelligence Can Help Better Manage the COVID-19 Pandemic. Int. J. Environ. Res. Public Health 2020, 17, 3176. [Google Scholar] [CrossRef]
- Scopus. Analyze Search Results, TITLE-ABS-KEY (“COVID-19” or “CoViD-19” or “Hubei Pneumonia” or “Novel Corona Virus”). Available online: https://08105qd69-1104-y-https-www-scopus-com.mplbci.ekb.eg/term/analyzer.uri?sid=38aae2fabdccac31270f9b430acb9525&origin=resultslist&src=s&s=TITLE-ABS-KEY%28%22COVID-19%22+OR+%22CoViD-19%22+OR+%22Hubei+Pneumonia%22+OR+%22Novel+corona+virus%22%29&sort=plf-f&sdt=b&sot=b&sl=84&count=8069&analyzeResults=Analyze+results&txGid=fc5eb31e2f64a965b06d73db82b71d2e (accessed on 23 May 2020).
- Wang, W.; Tang, J.; Wei, F. Updated understanding of the outbreak of 2019 novel coronavirus (2019-nCoV) in Wuhan, China. J. Med Virol. 2020, 92, 441–447. [Google Scholar] [CrossRef] [Green Version]
- Garnier-Crussard, A.; Forestier, E.; Gilbert, T.; Krolak-Salmon, P. Novel Coronavirus (COVID-19) Epidemic: What Are the Risks for Older Patients? J. Am. Geriatr. Soc. 2020, 68, 939–940. [Google Scholar] [CrossRef] [Green Version]
- Wei, M.; Yuan, J.; Liu, Y.; Fu, T.; Yu, X.; Zhang, Z.-J. Novel Coronavirus Infection in Hospitalized Infants Under 1 Year of Age in China. JAMA 2020, 323, 1313–1314. [Google Scholar] [CrossRef]
- Ivorra, B.; Ferrández, M.R.; Vela-Pérez, M.; Ramos, A.M. Mathematical modeling of the spread of the coronavirus disease 2019 (COVID-19) taking into account the undetected infections. The case of China. Commun. Nonlinear. Sci. Numer. Simul. 2020, 88, 105303. [Google Scholar] [CrossRef]
- Shen, C.; Chen, A.; Luo, C.; Zhang, J.; Feng, B.; Liao, W. Using Reports of Symptoms and Diagnoses on Social Media to Predict COVID-19 Case Counts in Mainland China: Observational Infoveillance Study. J. Med. Internet Res. 2020, 22, e19421. [Google Scholar] [CrossRef]
- Ayyoubzadeh, S.M.; Ayyoubzadeh, S.M.; Zahedi, H.; Ahmadi, M.; Niakan Kalhori, R.S. Predicting COVID-19 Incidence Through Analysis of Google Trends Data in Iran: Data Mining and Deep Learning Pilot Study. JMIR Public Health Surveill. 2020, 6, e18828. [Google Scholar] [CrossRef]
- Hashim, H.A.; El-Ferik, S.; Lewis, F.L. Neuro-adaptive cooperative tracking control with prescribed performance of unknown higher-order nonlinear multi-agent systems. Int. J. Control 2019, 92, 445–460. [Google Scholar] [CrossRef]
- Hashim, H.A.; El-Ferik, S.; Lewis, F.L. Adaptive synchronisation of unknown nonlinear networked systems with prescribed performance. Int. J. Syst. Sci. 2017, 48, 885–898. [Google Scholar] [CrossRef] [Green Version]
- Araújo, R.d.A.; Nedjah, N.; Oliveira, A.L.I.; Meira, S.R.d.L. A deep increasing–decreasing-linear neural network for financial time series prediction. Neurocomputing 2019, 347, 59–81. [Google Scholar] [CrossRef]
- Eltoukhy, A.E.E.; Wang, Z.X.; Chan, F.T.S.; Fu, X. Data analytics in managing aircraft routing and maintenance staffing with price competition by a Stackelberg-Nash game model. Transp. Res. Part E Logist. Transp. Rev. 2019, 122, 143–168. [Google Scholar] [CrossRef]
- Eltoukhy, A.E.E.; Wang, Z.X.; Chan, F.T.S.; Chung, S.H.; Ma, H.; Wang, X.P. Robust Aircraft Maintenance Routing Problem Using a Turn-Around Time Reduction Approach. IEEE Trans. Syst. Man Cybern. Syst. 2019, 1–14. [Google Scholar] [CrossRef]
- Chung, S.H.; Ma, H.L.; Chan, H.K. Cascading Delay Risk of Airline Workforce Deployments with Crew Pairing and Schedule Optimization. Risk Anal. 2017, 37, 1443–1458. [Google Scholar] [CrossRef]
- Misiunas, N.; Oztekin, A.; Chen, Y.; Chandra, K. DEANN: A healthcare analytic methodology of data envelopment analysis and artificial neural networks for the prediction of organ recipient functional status. Omega 2016, 58, 46–54. [Google Scholar] [CrossRef]
- Tsai, T.-H.; Lee, C.-K.; Wei, C.-H. Neural network based temporal feature models for short-term railway passenger demand forecasting. Expert Syst. Appl. 2009, 36, 3728–3736. [Google Scholar] [CrossRef]
- Magesh, S.; Niveditha, V.R.; Rajakumar, P.S.; RamMohan, S.R.; Natrayan, L. Pervasive computing in the context of COVID-19 prediction with AI-based algorithms. Int. J. Pervasive Comput. Commun. 2020, in press. [Google Scholar]
- Tuli, S.; Tuli, S.; Tuli, R.; Gill, S.S. Predicting the growth and trend of COVID-19 pandemic using machine learning and cloud computing. Internet Things 2020, 11, 100222. [Google Scholar] [CrossRef]
- WorldOmeter. COVID-19 CORONAVIRUS PANDEMIC. Available online: https://www.worldometers.info/coronavirus/ (accessed on 6 May 2020).
- Our World in Data. The COVID-19 Pandemic Slide Deck. Available online: https://ourworldindata.org/the-covid-19-pandemic-slide-deck (accessed on 30 April 2020).
- World Bank. World Bank Open Data. Available online: https://data.worldbank.org/ (accessed on 7 May 2020).
- United Nations Development Programme. Human Development Reports. Available online: http://hdr.undp.org/en/data (accessed on 8 May 2020).
- Agency, T.N. Country of Four Seasons: Iran A World inside A Country. Available online: https://www.tasnimnews.com/en/news/2016/10/03/1202507/country-of-four-seasons-iran-a-world-inside-a-country (accessed on 20 August 2020).
- Menezes, J.M.P.; Barreto, G.A. Long-term time series prediction with the NARX network: An empirical evaluation. Neurocomputing 2008, 71, 3335–3343. [Google Scholar] [CrossRef]
- Xie, H.; Tang, H.; Liao, Y.-H. Time series prediction based on NARX neural networks: An advanced approach. In Proceedings of the 2009 International Conference on Machine Learning and Cybernetics, Baoding, China, 12–15 July 2009; pp. 1275–1279. [Google Scholar]
- Boussaada, Z.; Curea, O.; Remaci, A.; Camblong, H.; Mrabet Bellaaj, N. A nonlinear autoregressive exogenous (NARX) neural network model for the prediction of the daily direct solar radiation. Energies 2018, 11, 620. [Google Scholar] [CrossRef] [Green Version]
- Marcjasz, G.; Uniejewski, B.; Weron, R. On the importance of the long-term seasonal component in day-ahead electricity price forecasting with NARX neural networks. Int. J. Forecast. 2019, 35, 1520–1532. [Google Scholar] [CrossRef]
- Rai, A.; Upadhyay, S. The use of MD-CUMSUM and NARX neural network for anticipating the remaining useful life of bearings. Measurement 2017, 111, 397–410. [Google Scholar] [CrossRef]
- Eltoukhy, A.E.E.; Wang, Z.X.; Chan, F.T.S.; Chung, S.H. Joint optimization using a leader–follower Stackelberg game for coordinated configuration of stochastic operational aircraft maintenance routing and maintenance staffing. Comput. Ind. Eng. 2018, 125, 46–68. [Google Scholar] [CrossRef]
- Plaehn, D.; Horne, J. A regression-based approach for testing significance of “just-about-right” variable penalties. Food Qual. Prefer. 2008, 19, 21–32. [Google Scholar] [CrossRef]
- Kim, Y.-S.; Yum, B.-J. Robust design of multilayer feedforward neural networks: An experimental approach. Eng. Appl. Artif. Intell. 2004, 17, 249–263. [Google Scholar] [CrossRef]
- Packianather, M.S.; Drake, P.R.; Rowlands, H. Optimizing the parameters of multilayered feedforward neural networks through Taguchi design of experiments. Qual. Reliab. Eng. Int. 2000, 16, 461–473. [Google Scholar] [CrossRef]
- Tabrizi, S.; Ghodsypour, S.H.; Ahmadi, A. Modelling three-echelon warm-water fish supply chain: A bi-level optimization approach under Nash–Cournot equilibrium. Appl. Soft Comput. 2018, 71, 1035–1053. [Google Scholar] [CrossRef]
- Menebo, M.M. Temperature and precipitation associate with Covid-19 new daily cases: A correlation study between weather and Covid-19 pandemic in Oslo, Norway. Sci. Total Environ. 2020, 737, 139659. [Google Scholar] [CrossRef]
- Şahin, M. Impact of weather on COVID-19 pandemic in Turkey. Sci. Total Environ. 2020, 728, 138810. [Google Scholar] [CrossRef]
- Tosepu, R.; Gunawan, J.; Effendy, D.S.; Ahmad, L.O.A.I.; Lestari, H.; Bahar, H.; Asfian, P. Correlation between weather and Covid-19 pandemic in Jakarta, Indonesia. Sci. Total Environ. 2020, 725, 138436. [Google Scholar] [CrossRef]
- Eltoukhy, A.E.E.; Chan, F.T.S.; Chung, S.H.; Niu, B. A model with a solution algorithm for the operational aircraft maintenance routing problem. Comput. Ind. Eng. 2018, 120, 346–359. [Google Scholar] [CrossRef]
- Eltoukhy, A.E.E.; Chan, F.T.S.; Chung, S.H.; Niu, B.; Wang, X.P. Heuristic approaches for operational aircraft maintenance routing problem with maximum flying hours and man-power availability considerations. Ind. Manag. Data Syst. 2017, 117, 2142–2170. [Google Scholar] [CrossRef]
- Milman, O. Coronavirus in America: Why the US has struggled to tackle a growing crisis. The Guardian. 24 March 2020. Available online: https://www.theguardian.com/us-news/2020/mar/24/coronavirus-america-donald-trump-crisis (accessed on 7 July 2020).
- BBC NEWS. Coronavirus: Brazil Records third-highest Covid-19 Infection Level. Available online: https://www.bbc.com/news/world-latin-america-52719391# (accessed on 7 July 2020).
- Collyns, D. Peru’s coronavirus response was ‘right on time’—So why isn’t it working? The Guardian. 20 May 2020. Available online: https://www.theguardian.com/global-development/2020/may/20/peru-coronavirus-lockdown-new-cases (accessed on 7 July 2020).
- Npr News. COVID-19 Numbers Are Bad in Ecuador. The President Says The Real Story Is Even Worse. Available online: https://www.npr.org/sections/goatsandsoda/2020/04/20/838746457/covid-19-numbers-are-bad-in-ecuador-the-president-says-the-real-story-is-even-wo (accessed on 8 June 2020).
Continent | Country |
---|---|
Europe | Spain, Italy, UK, Russia, and France |
North and South America | USA, Brazil, Canada, Peru, and Ecuador |
Asia | Turkey, Iran, China, India, and Saudi Arabia |
Africa | Egypt, South Africa, Morocco, Algeria, and Nigeria |
Hypothesis | p Value | Decision | Interpretation |
---|---|---|---|
Hypothesis # 1: population | 0.000 | Reject and pick | Population is significant |
Hypothesis # 2: median age index | 0.041 | Reject and pick | Median age index is significant |
Hypothesis # 3: public healthcare expenditure | 0.523 | Cannot Reject and reject | Public healthcare expenditure is not significant |
Hypothesis # 4: private healthcare expenditure | 0.000 | Reject and pick | Private healthcare expenditure is significant |
Hypothesis # 5: air quality as a CO2 trend | 0.476 | Cannot Reject and reject | Air quality as a CO2 trend is not significant |
Hypothesis # 6: number of arrivals in the countries/territories | 0.000 | Reject and pick | Number of arrivals in the countries/territories is significant |
Hypothesis # 7: education index | 0.047 | Reject and pick | Education index is significant |
Hypothesis # 8: seasonality as month of collecting data | 0.000 | Reject and pick | Seasonality as month of collecting data is significant |
Parameter | Level 1 | Level 2 | Level 3 |
---|---|---|---|
Learning rate | 0.01 | 0.1 | 0.3 |
Momentum | 0.1 | 0.3 | 0.5 |
Number of neurons in the first hidden layer | 1.5 × N = 336 | ||
Number of neurons in the second hidden layer | 0 | N/2 = 112 |
Continent | Country | NARX Neural Network-Based Algorithm | |||
---|---|---|---|---|---|
Root Mean Square Error (RMSE) | Spearman Correlation | Error Standard Deviation | |||
Correlation Factor | p-Value | ||||
Europe | Spain | 300 | 0.9687 | 0.000 | 902.74 |
Italy | 71 | 0.9725 | 0.000 | 215.38 | |
UK | 113 | 0.9770 | 0.000 | 343.13 | |
Russia | 150 | 0.9748 | 0.000 | 420.57 | |
France | 189 | 0.9753 | 0.000 | 587.11 | |
North and South America | USA | 786 | 0.9882 | 0.000 | 23,792.25 |
Brazil | 1146 | 0.9828 | 0.000 | 3463.52 | |
Canada | 54 | 0. 9566 | 0.000 | 163.93 | |
Peru | 148 | 0.9716 | 0.000 | 423.08 | |
Ecuador | 18 | 0.9712 | 0.000 | 57.03 | |
Asia | Turkey | 78 | 0.9753 | 0.000 | 251.77 |
Iran | 56 | 0.9784 | 0.000 | 170.81 | |
China | 14 | 0.9319 | 0.000 | 34.36 | |
India | 180 | 0.9946 | 0.000 | 550.63 | |
Saudi Arabia | 56 | 0.9720 | 0.000 | 171.79 | |
Africa | Egypt | 20 | 0.9696 | 0.000 | 61.55 |
South Africa | 79 | 0.9847 | 0.000 | 241.57 | |
Morocco | 28 | 0.9461 | 0.000 | 85.74 | |
Algeria | 7 | 0.9740 | 0.000 | 22.87 | |
Nigeria | 19 | 0.9695 | 0.000 | 57.61 |
Continent | Country | of NARX Neural Network-Based Algorithm | of Traditional Methods | Improvement of NARX over Traditional Method (%) |
---|---|---|---|---|
Europe | Spain | 300 | 49,492 | 99.39 |
Italy | 71 | 257 | 72.38 | |
UK | 113 | 988 | 88.56 | |
Russia | 150 | 371 | 59.59 | |
France | 189 | 1973 | 90.42 | |
North and South America | USA | 786 | 15,840 | 95.04 |
Brazil | 1146 | 3891 | 70.55 | |
Canada | 54 | 162 | 66.64 | |
Peru | 148 | 2229 | 93.36 | |
Ecuador | 18 | 777 | 97.68 | |
Asia | Turkey | 78 | 336 | 76.79 |
Iran | 56 | 152 | 63.10 | |
China | 14 | 334 | 95.81 | |
India | 180 | 615 | 70.73 | |
Saudi Arabia | 56 | 140 | 60.07 | |
Africa | Egypt | 20 | 32 | 37.07 |
South Africa | 79 | 149 | 46.98 | |
Morocco | 28 | 45 | 38.10 | |
Algeria | 7 | 12 | 39.78 | |
Nigeria | 19 | 27 | 28.76 |
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Eltoukhy, A.E.E.; Shaban, I.A.; Chan, F.T.S.; Abdel-Aal, M.A.M. Data Analytics for Predicting COVID-19 Cases in Top Affected Countries: Observations and Recommendations. Int. J. Environ. Res. Public Health 2020, 17, 7080. https://doi.org/10.3390/ijerph17197080
Eltoukhy AEE, Shaban IA, Chan FTS, Abdel-Aal MAM. Data Analytics for Predicting COVID-19 Cases in Top Affected Countries: Observations and Recommendations. International Journal of Environmental Research and Public Health. 2020; 17(19):7080. https://doi.org/10.3390/ijerph17197080
Chicago/Turabian StyleEltoukhy, Abdelrahman E. E., Ibrahim Abdelfadeel Shaban, Felix T. S. Chan, and Mohammad A. M. Abdel-Aal. 2020. "Data Analytics for Predicting COVID-19 Cases in Top Affected Countries: Observations and Recommendations" International Journal of Environmental Research and Public Health 17, no. 19: 7080. https://doi.org/10.3390/ijerph17197080