Static Learning in Prodigy
Static Learning in Prodigy
As opposed to learning through examination of completed search traces
as done in the EBL module, STATIC analyzes
domain knowledge to find control rules. To do this, it uses backward
chaining from fail states to find those operators whose application
lead to failure when attempting to reach some goal. Control rules are
then produced to avoid this failure during planning.
Return to the top of this architecture.
Current Location: Prodigy-Capabilities-Multi-Method Learning-Static Learning