Genetic Algorithms
Solving problems the natural way
Slide 2
Slide 2 text
Making Decisions - go to the market?
Slide 3
Slide 3 text
Making decisions - which car to buy?
Slide 4
Slide 4 text
Making decisions - mimic art?
wizardry!
50 semi transparent shapes Mona Lisa
Slide 5
Slide 5 text
Designing a car by mistake
Slide 6
Slide 6 text
Under the hood
Candidates Genome Fitness
List of candidate
solutions to the
problem.
“Genetic” code
for each
candidate that
describes the
characteristics
that will “evolve”
How to determine
the “fittest”
candidates
Slide 7
Slide 7 text
Under the hood
Candidates Genome Fitness
14aF2bdz12
9avg2bc1sd
44bg92jcks
120m
90m
1m
Slide 8
Slide 8 text
Under the hood
14aF2bdz12
shape
wheel size
wheel position
wheel density
chasis density
Slide 9
Slide 9 text
Under the hood
Candidates Genome Fitness
14aF2bdz12
14aG2bdz12
25aF2cdz12
?
?
?
Slide 10
Slide 10 text
Under the hood
Candidates Genome Fitness
14aF2bdz12
14aF2bc1td
?
?
?
9avg2bc1sd
Slide 11
Slide 11 text
Under the hood
Clone/mate to
build the next
generation
Find the fittest
candidate(s)
Randomly mutate
the genome
Build all
candidates
in generation
Slide 12
Slide 12 text
In the wild
Slide 13
Slide 13 text
In the wild
Slide 14
Slide 14 text
In the wild
- Flight / train scheduling
- Timetabling - e.g at a school
- Factory floor design
- Mechanical engineering
- Music production
- Financial modelling
- Code-breaking
- Artificial Intelligence
Slide 15
Slide 15 text
Genetic Algorithms
Solving problems the natural way