w2022-project-chrissy-ricky

w2022-project-chrissy-ricky created by GitHub Classroom

View on GitHub

Covid-19 & Making Predictions

Chrissy Aman & Ricky Sun

Summary

[Motivation & current literature]

Up until April 11, there are over 497,057,239 infections and 6,179,104 deaths since the beginning of the pandemic. Covid-19, or corona virus, has been impacting the world for over two years from all dimensions. Researchers have also appoached Covid-19 from all angles, and many of them seek to use currently available data to predict future trends.

Current research has already explored a wide variety of predictors for Covid-19 severity, infections, and deaths. Those research also evaluated potential impacts of the pandemic (such as economy and job market). Social media (such as Google search, Twitter twits), weather data (such as temperature, humidity), economic related data (such as GDP, interest rate), general health information (such as share of people aged over 70, share of smokers) are all factors that have been studied in current literature.

[Our research question and data]

In our research, we used data from “Our World in Data” Covid-19 public dataset, which includes data ranging from Covid-19 infections, gdp per capital, to vaccinations. Our project focus mainly on vaccintaions and death rate of Covid-19, but we also included many other variables and explored using multiple statistical tools.

[Methods]

We first did some preliminary analyses, such as maps, heat maps of correlations, scatter plots using death rate of Covid-19, vaccination rate and some other variables of our interest, such as Human Development Index (HDI), share of smokers, and so on.

We then move on to some regression analyses, starting from linear models. We first only include vaccination rate as x and death rate as y, and used data at more recent date at “2022-2-20”, which was the time that Omicron is the major variant virus in the world. Then, we include more variables in our model. We also included the same model but using a date during which Delta was the dominant variant (“2021-09-30”).

Because of the nature of panel data, the simple linear regressions selecting a specific day, the models may not be sufficient in making predictions. Also, when considering the time span from 2020 to 2022, the virus has mutated from alpha, Delta, to currently Omicron, and vaccinations have also been developed from first, second to booster doses. The simple linear regression may not be able to compromise those factors. Thus, we continued our analyses with more advanced manipulations.

Thus, We included regression discontinuity design and fixed effect model to further study the relationship between vaccination rate and death rate of Covid-19. We included ARiMA model and KNN machine learning model to predict total cases and death rates.

[Results]

According to our models, we found that: 1. Linear model: vaccination can negatively predict Covid-19 death 2. Linear model: stringency index (how strict a country’s policy is on Covid-19) can negatively predict Covid-19 death rate at a before Omicron date, but is positively predicting Covid-19 death rate at a Omicron date 3. Fixed effect: there is a clear negative correlation between vaccination rate and hospitalizations 4. RDD: the slope for regression of vaccination rate and Covid-19 death rate is positive until reaching the threshold of 50% and the slope turns negative after the 50% threshold 5. ARiMA: our ARiMA model gives a good forecast of future new cases 6. KNN: our KNN machine learning model categorize country into good and bad of coping with Covid-19 death rate with around 71% accuracy

There are many smaller findings, which are included in the presentation and rmd file.

[Limitations]

  1. Since we are using the global level by country data, the predictions and model accuracy may not be high if we want to predict local level Covid-19 cases or deaths
  2. We did not take policies into considerations, which is an important factor
  3. We have touched on Covid-19 variants, but our data does not actually distinguish between different infections
  4. Some of the data might need to be normalized or taken natural log to better deal with the screwness of the data

[Potential future studies]

  1. Use of county level data or more detailed datasets of certain country or region
  2. Compare effectiveness of different vaccines (Pfizer, Moderna, Sinovax, and so on)
  3. US and China can be a good natural experiment pair to compare with in terms of policies and Covid-19 data
  4. Use of social media data through machine learning models to make predictions

Presentation

Our presentation can be found here.

Data

Our dataset is coming from “Our World in Data” Covid-19 public data, together with data from JHU, WHO, CDC and World Bank. The data covers a wide range:

Hannah Ritchie, Edouard Mathieu, Lucas Rodés-Guirao, Cameron Appel, Charlie Giattino, Esteban Ortiz-Ospina, Joe Hasell, Bobbie Macdonald, Diana Beltekian and Max Roser (2020) - “Coronavirus Pandemic (COVID-19)”. Published online at OurWorldInData.org. viewed 25 February 2022, Retrieved from: ‘https://ourworldindata.org/coronavirus’ [Online Resource]

@article{owidcoronavirus, author = {Hannah Ritchie, Edouard Mathieu, Lucas Rodés-Guirao, Cameron Appel, Charlie Giattino, Esteban Ortiz-Ospina, Joe Hasell, Bobbie Macdonald, Diana Beltekian and Max Roser}, title = {Coronavirus Pandemic (COVID-19)}, journal = {Our World in Data}, year = {2020}, note = {https://ourworldindata.org/coronavirus} }

References

[1] Aboubakri, O., Ballester, J., Shoraka, H. R., Karamoozian, A., & Golchini, E. (2022). Ambient temperature and covid-19 transmission: An evidence from a region of Iran based on weather station and satellite data. Environmental Research, 209, 112887. https://doi.org/10.1016/j.envres.2022.112887

[2]Azuma, K., Kagi, N., Kim, H., & Hayashi, M. (2020). Impact of climate and ambient air pollution on the epidemic growth during COVID-19 outbreak in Japan. Environmental Research, 190, 110042. https://doi.org/10.1016/j.envres.2020.110042

[3] Chew, A. W., Pan, Y., Wang, Y., & Zhang, L. (2021). Hybrid deep learning of social media big data for predicting the evolution of COVID-19 transmission. Knowledge-Based Systems, 233, 107417. https://doi.org/10.1016/j.knosys.2021.107417

[4] Chiodini, I., Gatti, D., Soranna, D., Merlotti, D., Mingiano, C., Fassio, A., Adami, G., Falchetti, A., Eller-Vainicher, C., Rossini, M., Persani, L., Zambon, A., & Gennari, L. (2021). Vitamin D status and SARS-COV-2 infection and COVID-19 clinical outcomes. Frontiers in Public Health, 9. https://doi.org/10.3389/fpubh.2021.736665

[5] Fernández-González, R., Pérez-Pérez, M., Hervés-Estévez, J., & Garza-Gil, M. D. (2022). Socio-economic impact of covid-19 on the fishing sector: A case study of a region highly dependent on fishing in Spain. Ocean & Coastal Management, 221, 106131. https://doi.org/10.1016/j.ocecoaman.2022.106131

[6] Fritz, C., Dorigatti, E., & Rügamer, D. (2022). Combining graph neural networks and spatio-temporal disease models to improve the prediction of weekly COVID-19 cases in Germany. Scientific reports, 12(1), 3930. https://doi.org/10.1038/s41598-022-07757-5

[7] Hannah Ritchie, Edouard Mathieu, Lucas Rodés-Guirao, Cameron Appel, Charlie Giattino, Esteban Ortiz-Ospina, Joe Hasell, Bobbie Macdonald, Diana Beltekian and Max Roser (2020) - “Coronavirus Pandemic (COVID-19)”. Published online at OurWorldInData.org. viewed 25 February 2022, Retrieved from: ‘https://ourworldindata.org/coronavirus’ [Online Resource]

[8] Hou, Z., Du, F., Jiang, H., Zhou, X., & Lin, L. (2020). Assessment of public attention, risk perception, emotional and behavioural responses to the COVID-19 outbreak: Social media surveillance in China. https://doi.org/10.1101/2020.03.14.20035956

[9] Khan, A. R., Abedin, S., & Khan, S. (2022). Interaction of temperature and relative humidity for growth of COVID-19 cases and death rates. Environmental Research Letters, 17(3), 034048. https://doi.org/10.1088/1748-9326/ac4cf2

[10] Izcovich, A., Ragusa, M. A., Tortosa, F., Lavena Marzio, M. A., Agnoletti, C., Bengolea, A., Ceirano, A., Espinosa, F., Saavedra, E., Sanguine, V., Tassara, A., Cid, C., Catalano, H. N., Agarwal, A., Foroutan, F., & Rada, G. (2020). Prognostic factors for severity and mortality in patients infected with COVID-19: A systematic review. PloS one, 15(11), e0241955. https://doi.org/10.1371/journal.pone.0241955

[11] Lee, C.-K., Jung, E.-K., Kang, S.-E., Petrick, J. F., & Park, Y.-N. (2022). Impact of perception of covid-19 on NPI, job satisfaction, and customer orientation: Highlighting three types of npis for the airline industry. Journal of Air Transport Management, 100, 102191. https://doi.org/10.1016/j.jairtraman.2022.102191

[12] Li, J., He, X., Yuan Yuan, Zhang, W., Li, X., Zhang, Y., Li, S., Guan, C., Gao, Z., & Dong, G. (2021). Meta-analysis investigating the relationship between clinical features, outcomes, and severity of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pneumonia. American journal of infection control, 49(1), 82–89. https://doi.org/10.1016/j.ajic.2020.06.008

[13] Mudatsir, M., Fajar, J. K., Wulandari, L., Soegiarto, G., Ilmawan, M., Purnamasari, Y., Mahdi, B. A., Jayanto, G. D., Suhendra, S., Setianingsih, Y. A., Hamdani, R., Suseno, D. A., Agustina, K., Naim, H. Y., Muchlas, M., Alluza, H. H., Rosida, N. A., Mayasari, M., Mustofa, M., … Harapan, H. (2021). Predictors of COVID-19 severity: A systematic review and meta-analysis. F1000Research, 9, 1107. https://doi.org/10.12688/f1000research.26186.2

[14] Muhammad, L.J., Algehyne, E.A., Usman, S.S. et al. Supervised Machine Learning Models for Prediction of COVID-19 Infection using Epidemiology Dataset. SN COMPUT. SCI. 2, 11 (2021). https://doi.org/10.1007/s42979-020-00394-7

[15] Qin, Lei, Qiang Sun, Yidan Wang, Ke-Fei Wu, Mingchih Chen, Ben-Chang Shia, and Szu-Yuan Wu. 2020. “Prediction of Number of Cases of 2019 Novel Coronavirus (COVID-19) Using Social Media Search Index” International Journal of Environmental Research and Public Health 17, no. 7: 2365. https://doi.org/10.3390/ijerph17072365

[16] Ray, R. L., Singh, V. P., Singh, S. K., Acharya, B. S., & He, Y. (2022). What is the impact of COVID-19 pandemic on global carbon emissions? Science of The Total Environment, 816, 151503. https://doi.org/10.1016/j.scitotenv.2021.151503

[17] Xu, L., Magar, R., & Barati Farimani, A. (2022). Forecasting covid-19 new cases using Deep Learning Methods. Computers in Biology and Medicine, 144, 105342. https://doi.org/10.1016/j.compbiomed.2022.105342

[18] Yousefinaghani, S., Dara, R., Mubareka, S., & Sharif, S. (2021). Prediction of covid-19 waves using social media and google search: A case study of the US and Canada. Frontiers in Public Health, 9. https://doi.org/10.3389/fpubh.2021.656635

[19] Zoabi, Y., Deri-Rozov, S. & Shomron, N. Machine learning-based prediction of COVID-19 diagnosis based on symptoms. npj Digit. Med. 4, 3 (2021). https://doi.org/10.1038/s41746-020-00372-6