Meta-Reasoning Architecture
Architecture
Philosophy and Methodological Assumptions
Description of the Architecture
Interpreter
Logic Frames
Behaviors
Monitors
Domain Knowledge
Behavior Stack
Agent Properties
Style of Control
Modular Organization
Symbolic World Model
Glass Box Approach
Declarative Representations
Global Representation and Uniform Access to Knowledge
Efficiency of Knowledge Access
Representational Consistency (may be inconsistent)
Homogeneous (Uniform) Knowledge Representation
Frame-like Knowledge Representation
Meta-knowledge
Deliberative Learning
Non-monotonic Learning
Means-End Analysis Technique
Sequential Behavior
Coherence
Adequacy
Reflexive Response to Stimuli (Reactive Response)
Sensing Strategies and Attention
Malleable Architecture
Fixed Expressivity
List of Agent Properties for All Architectures
Capabilities
Multi-Method Learning
Planning and Situated Action
Replanning
Self Reflection
Meta-Reasoning
Focused Behavior and Processing/Selective Attention
Responding Intelligently to Interrupts and Failures
Lacks Parallel Behavior
Inability to Reason about Continuous Behavior
List of Capabilities for All Architectures
Environmental Considerations
Dynamic Environments
Simulated Environments
Complex Environments
Varying Priority
List of Environmental Considerations for All Architectures
Issues
References
Table of Contents
Current Location: Meta-Reasoning Architecture - Top