Sensing in Adaptive Intelligent Systems Architectures
Sensing in Adaptive Intelligent Systems Architectures
Perceptual input to an AIS undergoes considerable processing before it
is actually used by the reasoning system. The output of the preprocessors, which actually handle
the perceptors, is subjected to perceptual filters. It is the nature of
the filters applied that determines the system's sensitivity to
erroneous data. The AIS environment usually presents vast amounts of data, so errors in
individual readings are expected. The filters should not pick
out one reading as a sample for a time period but should instead
perform some more sophisticated amalgamation procedure. The more
knowledge about the types of errors present in the domain, the more
accurate the filtered data can be.
Return to the top of this architecture.
Go to a discussion of this capability
for multiple architectures.
Current Location: Adaptive Intelligent Systems-Capabilities-Sensing