ComputingComputer ScienceArtificial IntelligenceCognitive PsychologyEvolutionary PsychologyAgent-based ModelingCognitive ArchitectureCognitive EvolutionCognitive ModelingEvolutionary ComputationIntelligent AgentsArtificial AgentsCognitive AgentsCognitive Hierarchy ModelCognitive SystemsComplexityComputational ArchitectureReactive AgentsSymbolic Cognitive Modeling
This paper presents a generic model for a cognitive agent based on the hypothesis that the cognitive activity has three main characteristics: self-organization, evolutionary nature and history dependence. According to this model, a cognitive agent presents three levels: reactive, instinctive and cognitive. Each level, together with its lower levels, is intended to model a complete agent, each new level just increasing the behavior complexity. The generic model is instantiated into a computational architecture that integrates connectionist, evolutionary computation and symbolic approaches. In this architecture, the reactive level is an evolutionary environment where each organism is able to functionally link perceptive input -- represented by neural net outputs -- and effector control -- represented by the input to fuzzy control systems or Brooks' architectures. The instinctive level is based on an inductive learning process that optimizes the evolutionary process at the reactive level....
11
Genetic algorithms in search, optimization, and machine learning
Choice Reviews Online · 1989 · 49.3K citations
Circumscription—A form of non-monotonic reasoning
John McCarthy · Artificial Intelligence · 1980 · 2K citations
A. J. Kox, Martin J. Klein, Robert Schulmann · 1996 · 1.5K citations