Artificial Intelligence Expert Systems Genetic Algorithms

1
Artificial Intelligence / Expert SystemsGenetic
Algorithms

• Justin Gaudry
• July 3, 2007

2
Evolutionary Programming

• Survival of the fittest
• Optimization algorithms
• Simulate adaptive behavior in evolution
• Phenotypic

3
Genetic Algorithms

• Population
• Chromosome
• Allele
• Offspring

4
Genetic Algorithms

• Reproduction
• Crossover
• Mutation
• Fitness
• Culling
• Elitism
• http//www.codeproject.com/cs/algorithms/btl_ga.as
  p

5
Genetic Algorithms

• Selection Operators
• Random selection
• Proportional Selection
• Tournament selection
• Rank-based selection

6
Genetic Algorithms

• Represent the problem as a chromosome of a fixed
  length
• Choose the size of a population
• Crossover probability (high)
• Mutation probability (low)
• Define a fitness function
• Randomly generate an initial population
• Calculate the fitness of each

7
Genetic Algorithms

• Select a pair for mating
• Higher fitness first
• Create offspring
• Place offspring in population
• Repeat replacing less fit with more fit
  individuals
• Each iteration is a generation
• 15x x2

8
Fuzzy Systems

• Traditional set theory either part of a set or
  not
• True / false
• Fuzzy basis on uncertainty
• Approximate reasoning
• Element belongs to a set to a certain degree of
  certainty
• Reason with uncertain facts to infer new facts,
  each of which has a degree of certainty
• Modeling of common sense

9
Fuzzy Systems

• Nonstatistical uncertainty
• Based on vagueness, imprecision, and ambiguity
• Statistical uncertainty is based on laws of
  probabiity
• Control systems
• Gear transmission and braking systems
• Home appliances
• Traffic signals