J. Shifflet

Learn More
This paper presents a new technique, the hypothesis swarm problem-solving technique (HSPT), that allows swarms to operate on more complex problems in a general way that produces solutions faster than traditional swarms by emerging compatible sub-solutions cooperatively. Instead of agents simply reacting to their own stimuli, they produce hypotheses about(More)
Swarm algorithms rely on randomness to produce solutions for complex problems. The random selection of actions and chance interactions of agents force a swarm to attempt many behavioral possibilities - reinforcing the productive ones and dampening the dead ends. Randomness however, is a dual-edged sword: it is necessary to insure a wide range of agent(More)
  • 1