AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Is It All the Info or Just Some of It?
In conventional probability, we would imagine all the events that could happen and assign a number to them. So like we'd imagined, like it's just a list of numbers. And now you're saying, let's promote that from the start to a matrix,. To instead of n numbers and nd by n array of numbers. You have both the diagonal part of the matrix, which is your old fashioned set of probabilities, and more in fo is lurking there off diagonal and this. This lets you both compute the marginal probability, but then also keep some info about what was going on in the thing you are now ignoring.