Evolving Neural Network Architectures in a Computational Ecology

1
Evolving Neural Network Architectures in a
Computational Ecology

• Larry Yaeger
• Professor of Informatics, Indiana University
• Distinguished Scientist, Apple Computer
• Networks and Complex Systems
• Indiana University
• 18 October 2004

2
Wiring Diagram Learning Brain Maps
3
Motor Cortex Map
4
Motor Cortex Homunculus
5
Plasticity in Function
Orientation maps
Mriganka Sur, et al Science 1988, Nature 2000
6
Plasticity in Wiring
Patterns of long-range horizontal connections in
V1, normal A1, and rewired A1
Mriganka Sur, et al Nature 2000
7
Wiring Diagram Matters

• Relative consistency of brain maps across large
  populations
• Lesion/aphasia studies demonstrate very specific,
  limited effects
• Moderate stroke damage to occipital lobe can
  induce Charcot-Wilbrand syndrome (loss of dreams)
• Scarcity of tissue in localized portion of visual
  system (parietooccipital/intraparietal sulcus) is
  method of action for gene disorder, Williams
  Syndrome (lack of depth perception, inability to
  assemble parts into wholes)

8
Real Artificial Brain Maps
Distribution of orientation-selective cells in
visual cortex
9
Neuronal Cooperation
John Pearson, Gerald Edelman
10
Neuronal Competition
John Pearson, Gerald Edelman
11
The Story So Far

• Brain maps are good
• Brain maps are derived from
• General purpose learning mechanism
• Suitable wiring diagram
• Artificial neural networks capture key features
  of biological neural networks using
• Hebbian learning
• Suitable wiring diagram

12
How to Proceed?

• Design a suitable neural architecture
• Simple architectures are easy, but are limited to
  simple (but robust) behaviors
• W. Grey Walters Turtles
• First few Valentino Braitenberg Vehicles (1-3,
  of 14)
• Complex architectures are much more difficult!
• We know a lot about neural anatomy
• Theres a lot more we dont know
• It is being tried Steve Grands Lucy

13
How to Proceed?

• Evolve a suitable neural architecture
• It ought to work
• Valentino Braitenbergs Vehicles (4 and higher)
• We know it works
• Genetic Algorithms (computational realm)
• Natural Selection (biological realm)

14
Evolution is a Tautology

• That which survives, persists.
• That which reproduces, increases its numbers.
• Things change.
• Any little niche

15
Neural Architectures for Controlling Behavior
using Vision
16
(No Transcript)
17
(No Transcript)
18
What Polyworld Is

• An electronic primordial soup experiment
• Why do we get science, instead of ratatouille?
• Right ingredients in the right pot under the
  right conditions
• An attempt to approach artificial intelligence
  the way natural intelligence emerged
• Through the evolution of nervous systems in an
  ecology
• An opportunity to work our way up through the
  intelligence spectrum
• Tool for evolutionary biology, behavioral
  ecology, cognitive science

19
What Polyworld Is Not

• Fully open ended
• Even natural evolution is limited by physics (and
  previous successes)
• Accurate model of microbiology
• Accurate model of any particular ecology
• Though it is possible to model specific ecologies
• Accurate model of any particular organisms brain
• Though many neural models are possible
• A strong model of ontogeny

20
What is Mind?

• Hydraulics (Descartes)
• Marionettes (ancient Greeks)
• Pulleys and gears (Industrial Revolution)
• Telephone switchboard (1930s)
• Boolean logic (1940s)
• Digital computer (1960s)
• Hologram (1970s)
• Neural Networks (1980s - ?)
• Studying what mind is (the brain) instead of
  what mind is like

21
Polyworld Overview

• Computational ecology
• Organisms have genetic structure and evolve over
  time
• Organisms have simulated physiologies and
  metabolisms
• Organisms have neural network brains
• Arbitrary, evolved neural architectures
• Hebbian learning at synapses
• Organisms perceive their environment through
  vision
• Organisms primitive behaviors are neurally
  controlled
• Fitness is determined by Natural Selection alone
• Bootstrap online GA if required

22
(No Transcript)
23
Genetics Physiology Genes

• Size
• Strength
• Maximum speed
• Mutation rate
• Number of crossover points
• Lifespan
• Fraction of energy to offspring
• ID (mapped to bodys green color component)

24
Genetics Neurophysiology Genes

• of neurons for red component of vision
• of neurons for green component of vision
• of neurons for blue component of vision

• of internal neuronal groups

• of excitatory neurons per group
• of inhibitory neurons per group
• Initial bias of neurons per group
• Bias learning rate per group

• Connection density per pair of groups types
• Topological distortion per pair of groups types
• Learning rate per pair of groups types

25
Physiology and Metabolism

• Energy is expended by behavior neural activity
• Size and strength affect behavioral energy
  costs(and energy costs to opponent when
  attacking)
• Neural complexity affects mental energy costs
• Size affects maximum energy capacity
• Energy is replenished by eating food (or other
  organisms)
• Health energy is distinct from Food-Value energy
• Body is scaled by size and maximum speed

26
Perception Neural System Inputs

• Vision
• Internal energy store
• Random noise

27
Behavior Neural System Outputs

• Primitive behaviors controlled by single neuron
• Volition is level of activation of relevant
  neuron
• Move
• Turn
• Eat
• Mate (mapped to bodys blue color component)
• Fight (mapped to bodys red color component)
• Light
• Focus

28
Behavior Sample Eating
29
Behavior Sample Killing Eating
30
Behavior Sample Mating
31
Behavior Sample Lighting
32
Neural System Internal Units

• No prescribed function
• Neurons
• Synaptic connections

33
Evolving Neural Architectures
34
Neural System Learning and Dynamics

• Simple summing and squashing neuron model
• xi ? ajt sijt j
• ait1 1 / (1 e-xi)
• Hebbian learning
• sijt1 sijt ckl (ait1 - 0.5) (ajt - 0.5)

35
Emergent Species Joggers
36
Emergent Species Indolent Cannibals
37
Emergent Species Edge-runners
38
Emergent Species Dervishes
39
Emergent Behavior Visual Response
40
Emergent Behavior Fleeing Attack
41
Emergent Behaviors Foraging, Grazing, Swarming
42
A Few Observations

• Evolution of higher-order, ethological-level
  behaviors observed
• Selection for use of vision observed
• This approach to evolution of neural
  architectures generates a broad range of network
  designs

43
Is It Alive? Ask Farmer Belin

• Life is a pattern in spacetime, rather than a
  specific material object.
• Self-reproduction.
• Information storage of a self-representation.
• A metabolism.
• Functional interactions with the environment.
• Interdependence of parts.
• Stability under perturbations.
• The ability to evolve.

44
Information Is What Matters

• "Life is a pattern in spacetime, rather than a
  specific material object. - Farmer Belin
  (ALife II, 1990)
• Schrödinger speaks of life being characterized by
  and feeding on negative entropy (What Is Life?
  1944)
• Von Neumann describes brain activity in terms of
  information flow (The Computer and the Brain,
  Silliman Lectures, 1958)
• Informational functionalism
• Its the process, not the substrate
• What can information theory tell us about living,
  intelligent processes

45
Information and Complexity

• Chris Langtons lambda parameter (ALife II)
• Complexity length of transients
• ? rules leading to nonquiescent state /
  rules

High
Complexity
Low
0.0
1.0
?c
Lambda

• Crutchfield Similar results measuring
  complexity of finite state machines needed to
  recognize binary strings

46
Quantifying Life and Intelligence

• Measure state and compute complexity
• What complexity?
• Mutual Information
• Adamis physical complexity
• Gell-Mann Lloyds effective complexity
• What state?
• Chemical composition
• Electrical charge
• Aspects of behavior or structure
• Neuronal states
• Other issues
• Scale, normalization, sparse data

47
Future Directions

• Compute and record measure(s) of complexity
• Use best complexity measure(s) as fitness
  function
• More environmental interaction
• Pick up and put down pieces of food
• Pick up and put down pieces of barrier
• More complex environment
• More control over food growth patterns
• Additional senses
• More complex, temporal (evolved?) neural models

48
Future Directions

• Behavioral Ecology benchmarks
• Optimal foraging
• Patch depletion (Marginal Value Theorem)
• Patch selection (profitability vs. predation
  risk)
• Vancouver whale populations
• Evolutionary Biology problems
• Speciation (population isolation)
• Altruism (genetic similarity)
• Classical conditioning, intelligence assessment
  experiments

49
Future Directions

• Source code is available
• Original SGI version at ftp.apple.com in
  /research/neural/polyworld
• New Mac/Windows/X11 version coming soon, based
  on Qt
• Paper and other materials at rryy

50
Evolving Neural Network Architectures in a
Computational Ecology

• Larry Yaeger
• mailto larryy_at_indiana.edu
• http//pobox.com/larryy
• Networks and Complex Systems
• Indiana University
• 18 October 2004

51
But It Can't Be Done!

• "If an elderly but distinguished scientist says
  that something is possible he is almost certainly
  right, but if he says that it is impossible he is
  very probably wrong." - Arthur C. Clarke
• Humans are a perfect example of mind embedded in
  matter there is no point arguing about whether
  it is possible to embed mind in matter.
• The Earth is flat and at the center of the
  universe...

52
But Gödel Said So...

• No he didn't.
• Every consistent formalisation of number theory
  is incomplete.
• It is a huge leap to "AI is impossible".
• Indeed, the fact that human brains are capable of
  both expressing arithmetical relationships and
  contemplating "I am lying" bodes well for machine
  minds.
• The (formal) consistency of the human mind has
  most definitely not been proven.

53
Quantum Effects Required for Unpredictability

• With just three variables, Lorenz demonstrated
  chaotic, unpredictable systems.
• Even the 102 neurons and 103 synapses of
  Polyworld's organisms should provide adequate
  complexity.

54
Man Cannot Design Human Minds

• Even Gödel acknowledged that human-level minds
  might be evolved in machines.