- Scientists are repurposing a probability theory that is often illustrated by the ‘rationalization of reward’ behavior at slot machines
- The algorithm outperformed other processes currently being used to identify HIV hotspots
Scientists at Yale University recently published a report detailing their research into methods used to identify burgeoning HIV transmission zones.
The test of artificial intelligence used four different algorithms to study the accuracy of each selection method — one of which is commonly used to illustrate the rationalization of reward-seeking behavior in a user at a slot machine.
The lead author of the report, Gregg Gosalves is reported as saying: “When you walk into a casino and see a row of slot machines how do you decide which one to play and when it’s time to switch to another? What’s the best strategy to maximize your winnings? Mathematicians have created strategies called “bandit” algorithms to answer these questions – and we’ve used one of them as the basis for our approach to HIV testing.”
A synopsis of the study points out that four algorithms were simulated in over 250 tournament runs to determine which would adapt best to predict, track and trace HIV transmission zones.
Specifically, the algorithms were Thompson Sampling, Explore-then-exploit, Retrospection and a perfect, hypothetical Clairvoyance.
While the names sound like they might be for racehorses, they actually represent different decision making frameworks—and are therefore not the same “slot math” used to calculate the theoretical returns of a slot game.
In application, the Thompson Sampling logic appeared to identify 15% more cases than Explore-then-exploit and was consistent with Clairvoyance 90% of the time.
By optimizing the techniques and processes of HIV hotspot identification, scientists hope to be able to predict areas of high-risk and dispatch medical relief to these areas—saving money and lives.
Want to try an online casino?
Choose an approved casino from our carefully selected list. VIEW CASINOS