Algorithm for decision tree in continuous time
I'm attempting to build a fairly simple AI that has an input of 3 discreet
variables and is trying to predict a discreet output. This seems fairly
simple using an entropy calculation if you have a complete data set to
analyze beforehand. However, the program I am writing will be presented
one piece of data at a time and will have to learn to play a game as the
data is presented (eg. Playing a poker game or Rock Paper Scissors). How
would I be able to construct a tree on the fly as the data is received one
row at a time?
Thank you
No comments:
Post a Comment