Paper Info
Title
Predictive learning and cognitive momentum: a foundation for intelligent, autonomous systems
Abstract
An obvious characteristic of animal intelligence is the ability to recall learned or innate sequences. This phenomenon varies from such behavior as wasp nesting activity at the insect level, to alphabet memorization, spelling, music recitation, mathematical operation, motor action, rule application, event storage and recall, hypothesis formation, and script processing for humans. Kanerva (1988) has suggested much of this and proposed that activities such as these, and possibly much of what passes for intelligent behavior, can be explained as the result of a suitably implemented predictive sequence learning and recall system, and that such a system can be used to model intelligent behaviors in an automaton. Following Kanerva's lead, we introduce the idea of cognitive momentum, highlight the difference in stereotyped and non-stereotypical sequence processing, and discuss the potential ramifications of the latter with regard to realizing truly intelligent, autonomous capabilities. Finally, we report on steps toward implementation of a simulation model which can be used to demonstrate a number of the principles we address within a neurally inspired framework.
Year
DOI
Venue
1999
10.1145/306363.306392
ACM Southeast Regional Conference 2005
Keywords
Field
DocType
autonomous system,predictive learning,cognitive momentum,distributed systems,simulation model,sequence learning,development environment,object oriented,composition
Predictive learning,Object-oriented programming,Computer science,Development environment,Knowledge management,Theoretical computer science,Human–computer interaction,Autonomous system (Internet),Momentum,Cognition
Conference
ISBN
Citations 
PageRank 
1-58113-128-3
0
0.34
References 
Authors
11
1
Name
Order
Citations
PageRank
Steve Donaldson131.84