lean sluice
so for my thesis I will be making the AI for 2 games using the Q-Learning algorithm
those games are from a website named CodeCombat
below are the links for both games:
https://codecombat.com/play/ladder/frozen-fortress
https://codecombat.com/play/ladder/farmers-feud
all the game rules are already explained there when you enter the game so I am not going to explain it here
because Q-Learning is an ML algorithm and obviously needs training, it cannot be directly implemented into the game in real time, therefore I need to create a simulator for both games which does the training process
state and state space are 2 of the terms in Q-Learning
my question here is how to encode/represent the states and what is the size of the state space for both games? cuz from my point of view the state space is huge and it will not be possible to store all those states in the Q-Table cuz then my computer will run out of memory
so are there alternatives for a more optimal/efficient way to calculate the state space size and encode/represent the states for both games?
dapper cave