HereBoy:
A Fast Evolutionary
Algorithm
docsity.com
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Guidelines and tips
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Fast Evolutionary Algorithms - Embedded Intelligent Robotics - Lecture Slides, Slides of Robotics
Course title is Embedded Intelligent Robotics. This course is for Electrical engineering students. Though good thing is everyone can learn about robotics in this course. This lecture includes: Fast Evolutionary Algorithms, Claims, Evolutionary Time, Genetic Algorithm, Adaptive Mutation, Curve Optimization Experiment, Qualitative Comparison
Typology: Slides
2013/2014
1 / 16
This page cannot be seen from the preview
Don't miss anything!
Related documents
Partial preview of the text
Download Fast Evolutionary Algorithms - Embedded Intelligent Robotics - Lecture Slides and more Slides Robotics in PDF only on Docsity!
HereBoy:
A Fast Evolutionary
Algorithm
Claims
- Up to
100X faster
than a
Genetic
- Up to Algorithm
10X faster
than a
Simulated
- Significantly better Scalability Annealing
Genetic Algorithm
Population of
Individuals
Survival of the
Fittest Reproduction
with Crossover
Generation
n+
Generation
n
15
60
20 48
Population of
Individuals
Simulated Annealing
1-Bit
Mutate
Score
Chromosome
Mutated
Chromosome
Evaluate
If Better
If Worse
Probability
Test
You still keep
worse
solutions with
some
probability docsity.co
Adaptive Mutation
Convergence
Adaptive Mutation
Chromosome
Mutation Bits =
αααα
= MaxMutationRate = UserFraction x ChromosomeBits
MaxScore - MaxCurrentScore
MaxScore
How many bits to
mutate?
Adaptive Search
Convergence
Acceptance
Probability
0.
Acceptance
Probability
0.
Acceptance
Probability
0.
A B C D E F
Adaptive Search
A
B
Search Probability =
ρ
= MaxSearchProbability = UserFraction
MaxScore - MaxCurrentScore
MaxScore
How to select the probability value?
Curve Optimization Results
Iterations Solving Binary F
99% Complete
100% Complete
x
σ
σ σ
σ
x
σσσσ
Genetic
Algorithm
Annealing Simulated
HereBoy
Easy and Hard Binary F6 Experiments
HereBoy
Simulated Annealing
Genetic
Algorithm
Scale Between
Number of
iterations
docsity.co
Evolvable Hardware Experiment
Pattern Generation
Iteration = 0 Score = 800
Iteration = 3,
Score = 1,
Iteration = 65,
Score = 1,
Pattern Generator Circuit Statistics
10x
20x
Search Space Size
6400
25,
Probability of
Randomly Creating
100% Solution
Qualitative Comparison
HereBoy
Simulated Annealing
Genetic
Algorithm
Maximum Score