Beam local search - more states are retained (a population of states)
- evolutionary algorithms
- particle swarm optimisation
- ant colony optimisation
Swarm intelligence (collective intelligence)
A group of individuals that interact in order to achieve some objectives by collective adaptation to a global or local environment
Advantages:
- control is distributed among more individuals
- individuals local communicate
- system behaviour transcends the individual behaviour
- system is robust and can adapt to environment changes
Swarm
- more individuals, apparently non-organized, that are moving in order to form a group, but each individual seems to move in a particular direction
- inside the group can appear some social processes
- the collection is able to do complex tasks
- without a guide or an external control
- without a central coordination
- the collection can have performances better than the independent individuals
Collective adaptation → self-organisation
- set of dynamic mechanism that generates a global behaviour as a result of interaction among individual components
- rules that specify this interaction are executed based on local information only, without global references
- global behaviour is an emergent property of the system