Swarm Intelligence

Swarm intelligence (SI): the collective behaviour, self-organized systems
studies of the social behaviour of organisms in swarms inspired the development of very efficient algorithms
- e.g. ant colony optimization algorithms, particle swarm optimization algorithms, bacterial foraging, firefly, …

Ant-based algorithms

a population-based stochastic search → similar to GA
there is a population of ants, with each ant finding a solution then communicating with the other ants
- ants mark their path → attracts other ants
- when an ant reaches a fork in its path, it follows the way with the strongest smell (pheromones)
we can’t let pheromones build unbounded → evaporation is needed
only used for graph problems

WHILE (termination not satisfied) do
    CREATE ants
    FIND solutions
    UPDATE pheromones
    DAEMON activities (optional)
end WHILE

Transition probability: $P_{ij} (t) = \frac{T _{ij} ( t ) ^{α} ( \frac{1}{d _{ij}} ) ^{β}}{something goes here}$

Note

i give up i’ll finish these later (probably)