Download presentation
Presentation is loading. Please wait.
1
Hybridization of Search Meta-Heuristics Bob Buehler
2
A Recombination of Strengths Genetic Algorithm High correlation reproduction operators Fast computation excluding fitness Ant Colony Optimization Well suited in step- wise solution creation Strong local search using probabilistic pheromone model EAnt
4
The Power of Ants The World The Ant The Pheromone The Dream
5
Combinatorial Optimizers Ant Colony Optimization Traveling Salesman Problem S = The space of all possible solutions Τ = Pheromone model η = Heuristic values Step-wise solution creation About to select the next component for a partial solution c j = set of possible next components w(c i j ) = [τ i j ] α [η(c i j )] β p(c i j ) = w(c i j ) / Σ w(c j )
6
Basic ACO Algorithm Initialize pheromones and heuristics Iterate until termination condition Generate Solutions Update pheromones Decay all Increase those present in high fitness solutions
7
EAnt Evolving Pheromone Models Create random pheromone models as arrays of real values Let k ants walk the pheromone and create solutions Assign a fitness to the model equal to the average of all solutions created Use GA reproduction operators Profit
8
Testing EA vs ACO vs EAnt
9
Euclidean TSP 5 4 2 1 3 0 1435200 0X Y
10
EA Representation 1435200 14352100
11
EA Reproduction 3421500 1432500 1435200 1432500
12
EAnt Representation Pheromone Model is a two dimensional array M[n,m] where n is the node an ant is currently at and m is a node connected to n. Every element is initialized with a random value in the range [0,5).
13
EAnt Representation Example 4 2 1 3 0 5 41 12 03 31 1 2 45 4 2 3 0 31 1 01234 0 1 2 3 4 EAnt Genotype 143200 Environment
14
EAnt Reproduction Parameterized Uniform Crossover Gaussian Mutation with σ = 1
15
Results -Time Ranking 1. EA 2. ACO Step-wise cycle creation 3. EAnt Step-wise cycle creation O(n 2 ) individual size and reproduction
16
Results - EA and ACO Convergence
17
Results - EAnt Convergence (generations, individuals, fitness)
18
Hope
19
Final Thoughts Test for better final solution Different problem types EAnt pheromone model initialization 5 41 12 03 31 1 2 45 4 2 3 0 31 1 01234 0 1 2 3 4 5 50 15 05 20 2 1 25 2 2 1 0 10 1 01234 0 1 2 3 4 Improved?
20
Questions?
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.