Evolution and Artificial Life in Process Engineering

1
Evolution and Artificial Lifein Process
Engineering

• Janos Abonyi and Janos Madar

Department of Process EngineeringUniversity of
Veszprém, Hungary www.fmt.vein.hu/softcomp
2
(No Transcript)
3
Contents

• Neumanns self-replication scheme
• Self-reproduction and Artificial Life
• Stochastic optimization algorithms based
  Artificial Life
• Application examples

4
Introduction
Major works of John von Neumann

• Theory of Games and Economic Behavior
• Mathematical Foundations of Quantum Mechanics
• Computer and the Brain
• Theory of Self-reproducing Automata (posthumously)

Neumanns self-reproduction scheme can be
interpreted as the general version of stochastic
optimization techniques.
Neumann is called as the Father of game theory
and the Father of the modern computer and if
any person could be called the father of
Artificial life, it would be Neumann too.
5
Background of Self-Reproduction
In general, if we construct a machine, we use
even more complex machinery in its construction.
While engineering decreases the complexity, in
the nature the complexity may grow by evolution.
More complex
Less complex
Engineering
Evolution
Less complex
More complex
6
Neumanns Self-reproduction
The underlying idea of Neumanns
Self-reproduction is the separation of machines
from their descriptions.
The theoretical self-reproductive machine
consists of four parts
Universal constructor automation
Copyingmachine
Operativesystem
Descriptionof (ABC)
F(ABC)
A
B
C
7
Neumanns Self-reproduction
1. The C feeds the description into A
A
B
F(ABC)
C
8
Neumanns Self-reproduction
2. The A produce ABC
A
A
B
B
F(ABC)
C
C
9
Neumanns Self-reproduction
3. The C feeds the description into B
A
A
B
B
F(ABC)
C
C
10
Neumanns Self-reproduction
4. The B makes a copy of description
A
A
B
B
F(ABC)
F(ABC)
C
C
11
Neumanns Self-reproduction
5. Finally, we have a copy of the original system
A
A
B
B
F(ABC)
F(ABC)
C
C
12
Artificial Life
The logic model of self-reproduction views living
organisms as complex machineries.
Still we are not able to create such machineries,
but we can emulate them using computers.
Artificial Life

• A-life inspired stochastic optimization methods,
  e.g.
• Evolutionary Algorithms
• Particle Swarm Optimization

13
Game of life

• DEMO

14
Evolutionary computation
15
Motivations for EC

• Developing, analyzing, applying problem solving
  methods a.k.a. algorithms is a central theme in
  mathematics and computer science
• Time for thorough problem analysis decreases
• Complexity of problems to be solved increases
• Consequence
• Robust problem solving technology needed

16
The Main Evolutionary Computing Metaphor

• PROBLEM SOLVING
• Problem
• Candidate Solution
• Quality

• EVOLUTION
• Environment
• Individual
• Fitness

Fitness ? chances for survival and reproduction
Quality ? chance for seeding new solutions
17
Problem type 1 Optimisation

• We have a model of our system and seek inputs
  that give us a specified goal

• e.g.
• time tables for university, call center, or
  hospital
• design specifications, etc etc

18
Optimisation example Satellite structure
Optimised satellite designs for NASA to maximize
vibration isolation Evolving design
structures Fitness vibration resistance Evoluti
onary creativity
19
(No Transcript)
20
Problem types 2 Modelling

• We have corresponding sets of inputs outputs
  and seek model that delivers correct output for
  every known input

• Evolutionary machine learning

21
Modelling example loan applicant creditibility
British bank evolved creditability model to
predict loan paying behavior of new applicants
Evolving prediction models Fitness model
accuracy on historical data
22
Problem type 3 Simulation

• We have a given model and wish to know the
  outputs that arise under different input
  conditions

• Often used to answer what-if questions in
  evolving dynamic environments
• e.g. Evolutionary economics, Artificial Life

23
Simulation example evolving artificial societies

• Simulating trade, economic competition, etc. to
  calibrate models
• Use models to optimise strategies and policies
• Evolutionary economy
• Survival of the fittest is universal (big/small
  fish)

24
Brief History 1 the ancestors

• 1948, Turing
• proposes genetical or evolutionary search
• 1962, Bremermann
• optimization through evolution and recombination
  
• 1964, Rechenberg
• introduces evolution strategies
• 1965, L. Fogel, Owens and Walsh
• introduce evolutionary programming
• 1975, Holland
• introduces genetic algorithms
• 1992, Koza
• introduces genetic programming

25
Darwinian Evolution Survival of the fittest

• All environments have finite resources
• (i.e., can only support a limited number of
  individuals)
• Lifeforms have basic instinct/ lifecycles geared
  towards reproduction
• Therefore some kind of selection is inevitable
• Those individuals that compete for the resources
  most effectively have increased chance of
  reproduction
• Note fitness in natural evolution is a derived,
  secondary measure, i.e., we (humans) assign a
  high fitness to individuals with many offspring

26
The Genetic Algorithm

• Directed search algorithms based on the mechanics
  of biological evolution
• Developed by John Holland, University of Michigan
  (1970s)
• To understand the adaptive processes of natural
  systems
• To design artificial systems software that
  retains the robustness of natural systems

27
The Genetic Algorithm (cont.)

• Provide efficient, effective techniques for
  optimization and machine learning applications
• Widely-used today in business, scientific and
  engineering circles

28
Components of a GA

• A problem to solve, and ...
• Encoding technique (gene, chromosome)
• Initialization procedure
  (creation)
• Evaluation function (environment)
• Selection of parents (reproduction)
• Genetic operators (mutation, recombination)
• Parameter settings (practice and art)

29
Simple Genetic Algorithm

• initialize population
• evaluate population
• while TerminationCriteriaNotSatisfied
• select parents for reproduction
• perform recombination and mutation
• evaluate population

30
The GA Cycle of Reproduction
children
reproduction
modification
modified children
parents
evaluation
population
evaluated children
deleted members
discard
31
Population
population

• Chromosomes could be
• Bit strings
  (0101 ... 1100)
• Real numbers (43.2 -33.1 ...
  0.0 89.2)
• Permutations of element (E11 E3 E7 ... E1
  E15)
• Lists of rules (R1 R2 R3
  ... R22 R23)
• Program elements (genetic
  programming)
• ... any data structure ...

32
Reproduction
children
reproduction
parents
population
Parents are selected at random with selection
chances biased in relation to chromosome
evaluations.
33
Chromosome Modification
children

• Modifications are stochastically triggered
• Operator types are
• Mutation
• Crossover (recombination)

modification
modified children
34
Mutation Local Modification
Before (1 0 1 1 0 1 1 0) After (0
1 1 0 0 1 1 0) Before (1.38 -69.4
326.44 0.1) After (1.38 -67.5 326.44
0.1)

• Causes movement in the search space(local or
  global)
• Restores lost information to the population

35
Crossover Recombination

• P1 (0 1 1 0 1 0 0 0) (0 1 0 0 1 0 0
  0) C1
• P2 (1 1 0 1 1 0 1 0) (1 1 1 1 1 0 1
  0) C2
• Crossover is a critical feature of genetic
• algorithms
• It greatly accelerates search early in evolution
  of a population
• It leads to effective combination of schemata
  (subsolutions on different chromosomes)

36
Evaluation

• The evaluator decodes a chromosome and assigns it
  a fitness measure
• The evaluator is the only link between a
  classical GA and the problem it is solving

modified children
evaluated children
evaluation
37
Deletion
population

• Generational GAentire populations replaced with
  each iteration
• Steady-state GAa few members replaced each
  generation

discarded members
discard
38
Evolutionary Algorithms
Evolutionary Algorithm (EA) is an optimization
method which uses the computational model of
natural selection.
Initial population(first generation)
The EAs handle several potential solutions
simultaneously (population). These potential
solutions referred to as individuals. Every
individual represents a point in the search
space. Likes in the natural world, the more
successful individuals, which have bigger fitness
values, have higher probability to transmit their
genes into the new generation. The next
generation of individuals is generated from the
actual population using selection, mutation,
crossover.
Calculate fitness values(based on objectives)
Selection
Crossover and Mutation
New generation
End
39
An Abstract Example
Distribution of Individuals in Generation 0
Distribution of Individuals in Generation N
Genetic Optimization Toolbox gademox
40
A Simple Example

• The Traveling Salesman Problem
• Find a tour of a given set of cities so that
• each city is visited only once
• the total distance traveled is minimized

41
TSP Example 30 Cities
42
Representation

• Representation is an ordered list of city
• numbers known as an order-based GA.
• 1) London 3) Dunedin 5) Beijing 7)
  Tokyo
• 2) Venice 4) Singapore 6) Phoenix 8)
  Victoria
• CityList1 (3 5 7 2 1 6 4 8)
• CityList2 (2 5 7 6 8 1 3 4)

43
Crossover

• Crossover combines inversion and
• recombination
• Parent1 (3 5 7 2 1 6 4 8)
• Parent2 (2 5 7 6 8 1 3 4)
• Child (5 8 7 2 1 6 3 4)
• This operator is called the Order1 crossover.

44
Mutation

• Mutation involves reordering of the list
• 
  
• Before (5 8 7 2 1 6 3 4)
• After (5 8 6 2 1 7 3 4)

45
Solution i (Distance 941)
46
Solution j(Distance 800)
47
Solution k(Distance 652)
48
Best Solution (Distance 420)
49
Overview of Performance
50
A Simple Example

• The Gene is by far the most sophisticated
  program around.
• - Bill Gates, Business Week, June 27, 1994

51
Considering the EA Technology
Almost eight years ago ... people at Microsoft
wrote a program that uses some genetic things
for finding short code sequences. Windows 2.0 and
3.2, NT, and almost all Microsoft applications
products have shipped with pieces of code created
by that system. - Nathan Myhrvold, Microsoft
Advanced Technology Group, Wired, September 1995
52
Issues for EA Practitioners

• Choosing basic implementation issues
• representation
• population size, mutation rate, ...
• selection, deletion policies
• crossover, mutation operators
• Termination Criteria
• Performance, scalability
• Solution is only as good as the evaluation
  function (often hardest part)

53
Benefits of Evolutionary Algorithms

• Concept is easy to understand
• Modular, separate from application
• Supports multi-objective optimization
• Good for noisy environments
• Always an answer answer gets better with time
• Inherently parallel easily distributed

54
Benefits of Evolutionary Algorithms (cont.)

• Many ways to speed up and improve a GA-based
  application as knowledge about problem domain is
  gained
• Easy to exploit previous or alternate solutions
• Flexible building blocks for hybrid applications
• Substantial history and range of use

55
When to Use a EA

• Alternate solutions are too slow or overly
  complicated
• Need an exploratory tool to examine new
  approaches
• Problem is similar to one that has already been
  successfully solved by using a GA
• Want to hybridize with an existing solution
• Benefits of the GA technology meet key problem
  requirements

56
Some EA Application Types
57
Conclusions
The underlying idea of these algorithms is the
separation of solutions from descriptions of the
solutions. These algorithms work on these
descriptions and not on the solutions themselves.
Neumann proposed that non-trivial
self-reproduction should include the ability to
undergo inheritable mutations as well as the
ability to make another organism like the
original.
PSO and EA
Neumanns Self-reproduction
Self-replicating machines Potential
solutions Description of machines Coded
individuals Mutation of description Mutation
and recombination Growing complexity Growing
fitness
Self-organization scheme Selective organization
scheme
58
Tools

• http//www.it.uom.gr/pdp/DigitalLib/EC/ec_soft.htm