ML & AI are being sought to help with the early detection and spread of the COVID-19 pandemic.

Pandemic

Chile

Reference: How digital technologies are helping to flatten the curve https://www.weforum.org/agenda/2020/04/through-digital-technology-we-re-better-equipped-than-ever-to-fight-a-global-pandemic/ via @wef

China

Reference: How digital technologies are helping to flatten the curve https://www.weforum.org/agenda/2020/04/through-digital-technology-we-re-better-equipped-than-ever-to-fight-a-global-pandemic/ via @wef

Today, we have 4G and 5G, cloud computing, and artificial intelligence. These tools are helping to prevent, diagnose, and treat the disease, while helping scientists look for a cure.

Europe

Reference: https://www.imveurope.com/news/ai-aids-doctors-assess-covid-19-ct-scans

Computer vision can review hundreds of scans in seconds with 80% accuracy, and an “AI + human doctor” review is six times faster than evaluating images manually.

Supervised Learning:

• Outcome : Indicator of Flu sypmtoms ror Covid-19 symptoms.
• Sample: Either data from labs or random sample of individuals across the USA.

Features :

Random Forest or its variant can be a ML procedure used to predict the probability of contracting COVID-19.

Unsupervised Learning :

ML groups patients/respondents based on features like symptoms, demographics, medical history, antibodies into clusters Distinguish those clusters into possible disease diagnostics topics

Ethics & Challenges with ML results

• 1. False Negatives
• 2. False Positives
• 3. Race/Gender Bias & fairness