Pandemic, Analytics and Politics | Information age

With the public tired of lockdowns, overcrowded hospitals, vaccine shortages and policymakers drowning in the chaos of a COVID pandemic, Spanish data scientist Dr Nuria Oliver knew math and big data could take better decisions than a human politician.

His team set out to build big data models to predict the impact of COVID-19 and cost-effective policy interventions to minimize deaths, economic devastation and the spread of the dreaded disease.

The model took on 12 different ‘dimensions’ – or interventions such as closing workplaces and schools, wearing masks, and educating – so that AI modeling could create the best intervention. policy taking into account hospital burdens, economic costs and testing capabilities.

“When you look at the 12 dimensions, there were different levels of activation which led to 7.8 million possible combinations,” says Dr. Oliver, who has published hundreds of research papers and developed more than 40 patents. for data-driven solutions.

“No politician could possibly understand these 7.8 million different combinations, but a machine can. A machine can find the policy that satisfies the result with respect to the given cost, so it always comes up with the optimal policy. It is about the development of evidence-based policies. “

AI helps humans

Dr Nuria Oliver and her team, ValenciaIA4COVID, have been declared the world winners of the US $ 500,000 X price, which allowed them to create models for 236 countries to help navigate COVID-19.

“Models can make decisions that are not based on political interests or instinct but on evidence of what the data is telling you,” says Dr Oliver.

“The system offers policies with the best cost-benefit trade-off, so if your hospitals are empty you might want to open up a bit, or if the hospitals are full you will likely need to implement more stringent interventions to contain the transmission.”

Dr Oliver’s complex models take different metrics into account, but the ultimate challenge in winning the award was getting an “artificial politician” – also called the prescriber – to recommend a specific answer in 236 different countries or regions.

“Australia has been very easy – you have zero cases and you’re doing it right,” she quipped. “It’s the advantage of Australia and New Zealand to be an island. But you will also need to vaccinate so you don’t get left behind. “

Mathematics is complex, the results were not

The coronavirus pandemic has created many firsts – it wasn’t just the first time the world was adapting Zoom meetings faster than you could say ‘did you wash your hands’, but also the first global pandemic that captured and shared daily data on COVID-19 cases across the world.

Dr Oliver was working with the president of the Spanish region of Valencia in December 2020 when the team started working on the X Prize competition, and believes they could not have done the job without the daily collaboration they she maintained at the highest levels of government.

Building an accurate prediction for COVID-19 cases came using official data on COVID-19 cases and the Monitoring the Oxford Government’s response to COVID-19 dataset as primary data sources.

The team built neural network-based computational epidemiological models to predict COVID-19 cases 30, 60, and 180 days into the future.

The models used the confirmed number of COVID-19 cases and the interventions implemented to contain the pandemic in each of the 236 countries or regions around the world since March 2020.

The team also had to create a “prescriber” of 12 possible interventions, each with three to five different “levels”.

These interventions included:

  1. Closure of schools
  2. Closure of workplaces
  3. Cancellation of public events
  4. Restrict public gatherings
  5. Closure of public transport
  6. Make people say home
  7. Restrict internal movements
  8. International travel controls
  9. Public information campaigns, such as education
  10. Testing policies
  11. Contact follow-up
  12. Face cover or mask wearing

Like all good data science, data quality varied so some countries had to be excluded due to a lack of reliable data. Details of the work are being published in scientific journals but are not yet available, although there are more details on This site.

Dr Oliver said closing workplaces so people can work from home was the most effective policy to increase or decrease the number of COVID-19 cases in 236 countries and regions, followed closely by educational interventions, international travel controls and restrictions on gatherings.

Better modeling

The team was able to refine their models as Spain embarked on its third wave of COVID-19, bringing the death toll in Spain to around 77,000 in April 2021 – just as it finished the predictive model of the X Prize.

“We were able to accurately predict the course of the third wave of infections and test our predictions against real data,” says Dr Oliver.

Dr Oliver also led one of the greatest COVID response investigations, gain value qualitative insights in the pandemic and provide a voice for citizens.

This survey – which gathered 600,000 responses – revealed valuable information on the perception and impact of the pandemic on people’s lives, especially young people, as well as on the limitations of contact tracing.

“According to our survey, contact tracing apps are a complete failure in Spain, Italy and Germany,” she says.

How accurate was the perfect AI-constructed politician?

Given the hypothetical nature of the “prescriber” – or the so-called AI politician – the team could not assess their performance against the ground truth.

The predictor, however, provided policy recommendations in less than two hours, showing that politicians and governments could use these types of models to aid in complex day-to-day decision-making.

“I hope this will also help to advocate for more evidence-based political decision-making,” says Dr Oliver.

“Public administration and governments at all levels are not experiencing the digital transformation that most businesses have had to undergo over the past 15 years and the need to put in place a way to systematically collect data to inform policy in the future.”

Its greatest hope is that governments begin to standardize systematic data collection and analysis for social good.

“We can easily create a virtuous circle between data, people and technology so that public policy is informed by information,” she says.

Dr Nuria Oliver is Chief Data Scientist at Data-Pop Alliance, Chief Scientific Advisor at the Vodafone Institute and co-founder of ELLIS (The European Laboratory for Learning and Intelligent Systems) and co-founder of the Alicante ELLIS unit, dedicated researching “Human-Centered Artificial Intelligence (ity)”. She holds a PhD in Artificial Intelligence from the Massachusetts Institute of Technology (MIT) and has over 25 years of research experience in the areas of modeling and predicting human behavior from data and human-machine interaction.

Dr Nuria Oliver will host a webinar for the Institute of Analytics Professionals of Australia (IAPA) on her work with COVID-19 on Thursday, May 6 at 4 p.m. Click here to sign up.

Source link

About Robert Davis

Check Also

Italy targets unvaccinated with restrictions as cases rise | Government-and-politics

“We started to return to normality. We want to keep this normal, ”Draghi said at …

Leave a Reply

Your email address will not be published. Required fields are marked *