counterfactual regret minimisation
This a relatively new evaluation technique devised for poker bots to use when considering strategies. From NewScientist (20 August 2016, paywall) and Adam Kucharski:
“Regret” here refers to the difference between the expected pay-off of the action taken by their poker bot, named Cepheus, and the potential pay-off if it had acted differently. The technique involves Cepheus tweaking its strategy over the course of billions of hands, lowering its overall regret until it is as small as possible.
They are days when I wish I had taken up AI studies beyond that single course. It just sounds so interesting, even if I’m rather dubious about the uses of AI.
