Non Monotonic Learning in the Meta-Reasoning Architecture

Non Monotonic Learning in the Meta-Reasoning Architecture

MAX has no architecturally defined learning mechanism: any learning must be done by a user-supplied behavior. It was MAX's philosophy to give users the flexibility to create their own modules. In particular, the user has the option to do non-monotonic learning. This is when learning something new may invalidate knowledge that previously was considered true. It is often useful for handling default knowledge.


Return to the top of this architecture.

Go to a discussion of this property for multiple architectures.


Current Location: Meta-Reasoning - Properties - Non Monotonic Learning

Go to NEXT page.