AI Talk: Modeling inoculations and machine learning startups

January 22nd, 2021 / By V. “Juggy” Jagannathan, PhD

This week’s AI Talk…

Patient being vaccinated

Modeling inoculations

Recently we have been hearing this familiar refrain: “Vaccines do not prevent COVID-19– vaccinations do!”  It looks like we now have an approach to help figure out a timeline for COVID-19 vaccination. I saw this article in AI in Healthcare which referenced research on how to plan for vaccinations.

Researchers from York University in Toronto have developed a new tool that can help vaccination planners. What does it do? It helps planners estimate how many people can be vaccinated using drive-through vaccination centers customized to their locality. How does it do it? The researchers have developed a machine learning model that, when provided with parameters like number of lanes, number of shifts, pre-registration age and a host of other variables, predicts how many people can get vaccinated.

How researchers train the model is quite interesting. They have developed a detailed simulation model which they have used to simulate 125,000 trial runs so far. Next, researchers used the results of these simulations to train a neural network model. That model is quite small and performs just as well as the simulation model while requiring fewer resources to run. The model can be downloaded to an app and used by planners. Mass drive-through clinics must be in consideration for the new administration and perhaps this tool will help.

ML Startups

I saw a recent report in Forbes about roughly 10,000 startups that are all focused on machine learning! Who knew? Roughly $30 billion in overall funding goes to these startups. Forbes also spotlighted 25 startups they believe show promise. I don’t know what filters they used to select the 25 from 10,000, but the startups selected do appear interesting. These startups support platforms for machine learning and predictive data analytics for a range of applications:

  • Real-time monitoring of manufacturing lines
  • Non-expert users
  • R&D tax write-offs
  • Explainable solutions
  • Fraud detection
  • Fitness related
  • Marketing and sales support
  • Deep fake and fraud detection
  • Food industry related
  • Cybersecurity
  • Re-skilling the labor force
  • Population health
  • Motorcycle safety
  • Conversational assistants
  • Managing model versions
  • Vision and image focused

This list is proof that AI is permeating every aspect of our lives.

Conventional wisdom has it that one in ten startups succeed. If that holds, a mind-blowing 1,000 startups will succeed in this space! Many may be gobbled up by Big Tech, but regardless of who is at the helm, one thing seems certain: We are going to see a whole slew of solutions that we have never seen before. This is the Wild West of AI.

I am always looking for feedback and if you would like me to cover a story, please let me know. “See something, say something!” Leave me a comment below or ask a question on my blogger profile page.

V. “Juggy” Jagannathan, PhD, is Director of Research for 3M M*Modal and is an AI Evangelist with four decades of experience in AI and Computer Science research.