Difference between revisions of "ArtificialIntelligenceResearch"

Revision as of 12:48, 21 June 2015

Artificial Intelligence Research

@@Home -> ArtificialIntelligenceResearch

Artificial Intelligence research covers approaches of constructing intelligence feature which is somehow like biological one, by means of software program.

Terms

• artificial intelligence: a characteristic, enabling full replacement of a human for a specific subset of human functions

Research targets

See research status grouped by target

• Associative Memory
• Emotional Machine (suppressing mind areas instead of adding smth) ==
• Find whether Semantic Networks can help
• Evaluate reinforcement learning to allow learn state-to-action strategies
• Choose a model for low-level associative memory
• Evaluate Hierarchical Temporal Memory to create recognition engine
• Create a reasonable mind model, starting from draft in Vision
• Equivalence to biological brain.

external examples:

• link

• link

• link

• link

• link

• link

Interesting Things

• MFA controllers are ideal for environmental simulation and test control. Avoiding process models means lower development costs and faster time to market, more robust and precise control means better lab test results, and no manual tuning means lower operating costs and longer up time - see mfaoverview.html

Useful resources

docs:

 # Reinforcement Learning: Introduction
 # Journal: Developing Intelligence
 # AAAI: Congnitive Science
 # Hierarchical Temporal Memory, pdf

video:

 # HTM by Jeff Hawkins - Hierarchical Temporal Memory: Theory and Implementation
 # Way of thinking model - Emotion Machine: Commonsense Thinking, Artificial Intelligence, and the Future of the Human
 # Mind Components