Physics
Scientific paper
Jan 1996
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=1996adspr..18..221c&link_type=abstract
Advances in Space Research, Volume 18, Issue 11, p. 221-221.
Physics
Scientific paper
The objective of this paper is to suggest that introspective reasoning achieved through a two-parts architecture using Neural networks as a building block, could be an interesting ``intelligent tool'' for robotic systems. Particularly for those acting in dynamic and unstructured environments such as the surroundings of a space station. The first part learns and represents a known set of instructions. The second part memorizes it. When the agent has an unknown input, the second part gives a reasoned output doing introspection over its gained experience. This output is learned on-line by the first part and the process is repeated for each new input. This work combines concepts from Artificial Intelligence, Cognitive Science and Neurophysiology. The result is a system that is capable of capturing attentional processes and reasoning based on its experience. Conclusions on these processes, and the interactions between the two parts of the architecture are addressed.
No associations
LandOfFree
Robotic systems: intelligence through introspective reasoning does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Robotic systems: intelligence through introspective reasoning, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Robotic systems: intelligence through introspective reasoning will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-967751