27-YEAR-OLD CHRISTMAS BECOME THE EXACTLY ‘PRIOR’ OF COVID-19. Youyang Gu, a 27-year-old man living with his parents in California, USA, unexpectedly became an accurate predictor of the Covid-19 epidemic.
In early 2020, the whole world will rely on two reputable forecasting systems to assess the severity of the Covid-19 pandemic, according to Bloomberg.
Including a system developed by Imperial College London, UK. The other system is a product of the Institute of Health Measurement and Evaluation (IHME), based in Seattle, Washington state, USA.
But the two systems showed significantly different results. The British system warns the US could record 2 million Covid-19 deaths in the summer, while IHME forecasts that the figure will stop at only about 60,000.
In the end, neither system gives accurate results. As of early August 2020, the United States had recorded about 160,000 Covid-19 deaths.
The inaccurate prediction results attracted the attention of a young data scientist, Mr. Youyang Gu.
SUPER STAR DATA
Mr. Gu, 27 years old, holds a master’s degree in mathematics, engineering and computer science from the Massachusetts Institute of Technology (MIT). Although he has no medical training, he is confident that his data models can work in a pandemic.
In mid-April 2020, Gu spent a week building a tool to predict the death toll from the pandemic. He has also set up a website to report epidemiological data. Not long after, Gu’s algorithm gave accurate results, surpassing the prediction model from leading research institutes in the world.
“This is the only model that works,” says data expert Jeremy Howard from the University of San Francisco. “The other models have repeatedly become useless. But those who set them up do not consider and solve the problem, the press does not care much “.
“Everyone’s life depends on this,” concluded Mr. Howard. And Youyang Gu is the only one who is willing to review and process data properly.
In fact, Gu’s prediction model is quite simple. First, he considered the relationship between the number of Covid-19 tests and the number of hospitalizations. From there, he found that the states and federal governments that provided data were inconsistent.
Only the death toll is consistently updated, according to Gu. “Other models use many data sources,” he said. But I decided to rely on the number of deaths in the past to predict this figure in the future. This is the source of information with the least ‘noise’.
Although not perfect, Mr. Gu’s model has operated effectively from the moment it came into operation. By the end of April 2020, Gu predicts that the United States will record about 80,000 deaths from Covid-19 on May 9. Actual figures were 79,926 cases.
Gu also predicts that the deaths in the US on May 18 and May 27 will be 90,000 and 100,000, respectively. Again, the estimates match the reality.
Notably, Gu also predicted the second wave of epidemics, as soon as many states in the US removed their anti-epidemic measures. Meanwhile, the IHME
Institute believes that the corona virus will disappear thanks to social stretching measures.
IMPORTANT CONTRIBUTION
In March 2020, IHME faced some criticism for its inability to accurately predict epidemiological data. However, the unit is still funded by Bill & Melinda Gates Foundation with more than 500 million USD.
On April 19, 2020, Gu issued the warning of the second wave of epidemics. At the same time, former President Donald Trump cites the IHME prediction to show that the war against the corona virus will end soon.
Officials from IHME are also actively promoting their research results. Mr. Gu recalls: “IHME is constantly appearing in the media, trying to convince everyone that the number of deaths will reach zero by July”.
“I think this job is in vain,” he commented. Anyone can realize the death toll will remain between 1,000 and 1,500 cases over a period of time.
Director Christopher Murray of IHME, committed the forecast will be radically improved after April. But at this time, public opinion has begun to pay attention to Mr. Gu’s prediction model.
At the end of April 2020, the famous scientist Carl Bergstrom of the University of Washington posted about Gu’s model on Twitter. Not long after, the US Centers for Disease Control and Prevention (CDC) also published Gu’s figures on the official website.
As the pandemic continues, Gu regularly attends meetings with the CDC and professional statisticians and epidemiologists.
At the same time, traffic to Mr. Gu’s website exploded. Every day, millions more people sign up for accounts to update the progress of the disease in the area where they live.
Thanks to public interest, many predictive models began to appear during the summer of 2020. Associate Professor Nicholas Reich at the University of Massachusetts Department of Biological Statistics and Epidemiology collected about 50 models. predict and compare their accuracy. Mr. Reich commented: “Youyang’s model is always in the top most accurate”.
In November 2020, Gu decided to suspend his forecast model. Mr. Reich commented: “Youy
