Connor's Notes

⬅️ Back to portfolio

COSC 3P71: Lecture 10

2 min read

Announcements

  • Brock AI Day - November 6

Genetic Algorithms

Setting GA parameters

  • Parameters: selected according to the problem
    • how many individuals (chromosomes) will be in the population
      • too few: soon all chromosomes will have same traits & little crossover effect
      • too many: computation time will be expensive
    • mutation rate
      • too few: slow changes
      • too many: desired traits are not retained
    • how are individuals selected for mating? crossover points?
    • what should the probabilities of operations are used?
    • should a chromosome appear more than once in a problem?
    • fitness criteria
  • genetic algorithm can be computationally expensive → need to keep bounds on a GA parameters and GA analysis

Note

legendary ombuki rant about labs and robots

Evolutionary Algorithms

  • Recombination: crossover
# Pseudocode for typical evolutionary algorithm

BEGIN
    INITIALIZE population with random candidate solutions
    EVALUATE each candidate
    REPEAT UNTIL (TERMINATION CONDITION is satisfied) DO
        1. SELECT parents
        2. RECOMBINE pairs of parents
        3. MUTATE resulting offspring
        4. EVALUATE new candidate
        5. SELECT individuals for the next generation
    END
END

Note

read the rest of the slides on ur own (when/if they get posted)

  • small mutation rate (slides say 80%, go WAY lower)
  • soft time window: come around 4pm (3:45-4:15)
  • hard time window: come exactly at 4pm, bank closes at 5pm, etc

Graph View