Next: Decision Trees
Up: Delayed Reward
Previous: Delayed Reward
In many cases, what we would
like to do is partition the environment into regions of states that
can be considered the same for the purposes of learning and generating
actions. Without detailed prior knowledge of the environment, it is
very difficult to know what granularity or placement of partitions is
appropriate. This problem is overcome in methods that use adaptive
resolution; during the course of learning, a partition is constructed
that is appropriate to the environment.
Leslie Pack Kaelbling
Wed May 1 13:19:13 EDT 1996