UNDERSTANDING THE BRAIN

1
UNDERSTANDING THE BRAINS EMERGENT PROPERTIES

• Don Miner, Marc Pickett, and Marie desJardins
• Multi-Agent Planning Learning LabUniversity of
  Maryland, Baltimore County
• March 6, 2008
• don1_at_umbc.edu
• http//maple.cs.umbc.edu/don/projects/SAF

2
RULE ABSTRACTION

• Rule Abstraction the process of learning
  correlations between swarm-level properties and
  low-level parameter values.
• Enables control of swarms in terms of the
  swarm-level properties.
• Enables predictions of emergent behavior from the
  agent-level parameters.
• Results end-user control over swarms, swarm
  planners, richer applications.

3
RULE ABSTRACTION

• Low-level parameters
• Abstract property
• Mapping function
• Reverse mapping function
• The learning problem is defining
  and

4
(No Transcript)
5
RULE HIERARCHIES

• Natural extension to rule abstraction
• Abstract properties are used as low-level
  parameters of higher-level abstract properties.
• Changes anywhere in the hierarchy are propagated
  throughout.

6
THE MIND AS A RULE HIERARCHY?

• Looking at the mind as an emergent property of
  the brain, can we define a rule hierarchy that
  models intelligence?
• Can we learn how parameter values of atom-level
  programs influence emergent properties of the
  brain?
• Is there a way to break up the concept of
  intelligence into sub-emergent properties -- or
  does it all just emerge in one step?
• What other benefits are there of thinking of the
  brain as an emergent system?