ERE maintains several assumptions in its development:
- Plan execution is formalized as a form of closed-loop control, where a plan describes a desired behavior, and feedback from the environment is used to measure success or deviation.
- Actions can fail, so the results of actions must be validated in a fashion of truth maintenance.
- Learning within ERE assumes that any single body of knowledge is controvertible, and it is predicted that advantages may be gained over other systems that utilize strictly analytic methods that assume monotonic domain theory.
To return, press HOME.
Press NEXT for the next document.