liveBook · Manning

Neighborhood best

As explained in Optimization Algorithms: AI techniques for design, planning, and control problems

The best position among a restricted neighborhood containing only a few particles.

Neighborhood best (nbest) is the best position found by the particles in a selected neighborhood. When that neighborhood contains only a few particles, it is called the local best (lbest); when it contains the entire swarm, it is the global best (gbest). section 9.2

Use neighborhood best in particle swarm optimization to share useful information without requiring every particle to follow the entire swarm:

  • Evaluate each particle’s fitness and update its personal best when the current position improves on its previous best. section 9.2.2
  • Update each neighborhood’s best position, then use nbest or lbest in place of gbest when calculating particle motion. section 9.2.2 section 9.2.1
  • In synchronous PSO, particles use a common update stage; in asynchronous PSO, each particle can use the most recent neighborhood information during its update. section 9.2.2

Restricted versus whole-swarm neighborhoods

A particle swarm divided into small neighborhoods uses each neighborhood’s lbest to guide motion. Replacing those restricted neighborhoods with the whole swarm makes nbest identical to gbest. section 9.2 section 9.2.2

Figure 9.4 Synchronous and asynchronous PSO

Compare synchronous and asynchronous PSO update flow, including where the neighborhood-best update occurs. section 9.2.2

How Optimization Algorithms: AI techniques for design, planning, and control problems covers it

How deep it goes
Defines nbest and distinguishes lbest from gbest, then connects neighborhood updates to synchronous and asynchronous procedures.
Examples
Contrasts a few-particle neighborhood with the whole swarm and illustrates their update order through PSO flow figures.
What you can do after
Apply neighborhood best by replacing global-best guidance with local neighborhood information in particle motion updates.
How it connects
Connects personal best, global best, particle swarm optimization, motion equations, and synchronous or asynchronous updates.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Optimization Algorithms: AI techniques for design, planning, and control problems. The section links point to the parts of the book it draws on.