## The fully informed particle swarm: Simpler, maybe better (2004)

Venue: | IEEE Transactions on Evolutionary Computation |

Citations: | 68 - 3 self |

### BibTeX

@ARTICLE{Mendes04thefully,

author = {Rui Mendes and James Kennedy and José Neves},

title = {The fully informed particle swarm: Simpler, maybe better},

journal = {IEEE Transactions on Evolutionary Computation},

year = {2004},

volume = {8},

pages = {204--210}

}

### OpenURL

### Abstract

The canonical particle swarm algorithm is a new approach to optimization, drawing inspiration from group behavior and the establishment of social norms. It is gaining popularity, especially because of the speed of convergence and the fact it is easy to use. However, we feel that each individual is not simply influenced by the best performer among his neighbors. We thus decided to make the individuals “fully informed. ” The results are very promising, as informed individuals seem to find better solutions in all the benchmark functions.

### Citations

1157 |
M.: On computable numbers, with an application to the Entscheidungsproblem
- Turing
- 1936
(Show Context)
Citation Context ...ity regarding this question. How do we know if a function has a solution or not? Of course, we have known since Turing that we cannot tell with certainty whether an algorithm will ever reach finality =-=[9]-=-, that is in this case, whether a problem can be solved. But even though there is no certainty, there are clues. For instance, if it is believed that a cause and effect relationship exists among varia... |

528 |
Quasi-experimentation: Design and Analysis Issues for Field Settings
- Cook, Campbell
- 1979
(Show Context)
Citation Context ... the fitness landscape. Even if the causal relationship is noisy, or if the relationship involves variables not mentioned in the function (e.g., the “third variable problem” in correlational research =-=[10]-=-), it is often possible to find useful features on the function landscape. Another clue that a function might be solvable is when it is compressible. The easiest-to-spot form of this clue exists when ... |

447 | The particle swarm: explosion, stability and convergence in a multidimensional complex space
- Clerc, Kennedy
- 2009
(Show Context)
Citation Context ... position found by the particle, and � Pg refers to the position found by the member of its neighborhood that has had the best performance so far. The Type 1 ′′ constriction coefficient is often used =-=[1]-=-: �vt+1 = χ (�vt + � U[0, ϕ1] ⊗ ( � Pi − � Xt) + � U[0, ϕ2] ⊗ ( � Pg − � Xt)) (3) �Xt+1 = � Xt + �vt+1 The two versions are equivalent, but are simply implemented differently. The second form is used ... |

192 |
Parameter selection in particle swarm optimization
- Shi, Eberhart
- 1998
(Show Context)
Citation Context ...ation strategies, and algorithm details. The types of topologies have been described and are shown in Fig. 1. Two kinds of initialization strategies were used, which we called, after Shi and Eberhart =-=[12]-=- “symmetrical” and “asymmetrical”. Symmetrical initialization is performed over the entire spectrum of valid solutions, while asymmetrical initialization started particles with an offset, so they were... |

107 | R.Mendes, ”Population structure and Particle Swarm Performance
- Kennedy
- 2002
(Show Context)
Citation Context ...rangements of the neighborhoods can be thought of as social networks. It should be appreciated that the topological structure of the population controls its exploration versus exploitation tendencies =-=[3]-=-, [4]. (8)sIEEE TRANSACTIONS OF EVOLUTIONARY COMPUTATION, VOL. 1, NO. 1, JANUARY 2025 5 The behavior of each particle is affected by its local neighborhood, which can be seen as a single region in the... |

92 |
Small worlds and mega-minds: Effects of neighborhood topology on particle swarm performance
- Kennedy
(Show Context)
Citation Context ...ments of the neighborhoods can be thought of as social networks. It should be appreciated that the topological structure of the population controls its exploration versus exploitation tendencies [3], =-=[4]-=-. (8)sIEEE TRANSACTIONS OF EVOLUTIONARY COMPUTATION, VOL. 1, NO. 1, JANUARY 2025 5 The behavior of each particle is affected by its local neighborhood, which can be seen as a single region in the popu... |

57 |
An Off-The-Shelf PSO
- Carlisle, Dozier
- 2001
(Show Context)
Citation Context ... of the two terms keeps the particle searching between and beyond a region defined by the two points. While some investigators have looked at schemes for differentially weighting the two terms (e.g., =-=[2]-=-), the limits for the two uniform distributions are usually the same. That is, the total weight of ϕ is partitioned into two equal components. (6)sIEEE TRANSACTIONS OF EVOLUTIONARY COMPUTATION, VOL. 1... |

24 |
No Free Lunch Theorems for search. Santa Fe
- Wolpert, Macready
- 1995
(Show Context)
Citation Context ... some functions and not on others. What we hope for is a problem-solver that can work well with a wide range of problems. This line of reasoning drives us head-on into the No Free Lunch theorem (NFL) =-=[6]-=-, [7]. A. Free Lunch NFL asserts that no algorithm can be better than any other, over all possible functions. This seems to be true because of two classes of functions: deceptive ones, and random ones... |

13 | Optimization is easy and learning is hard in the typical function
- English
- 2000
(Show Context)
Citation Context ...ns, it must be true that gradients lead away from the optimum at least as often as they lead the searcher toward it. The second class, random functions, contains very many more members than the first =-=[8]-=-. In fact, when all possible functions are considered, it seems certain — indeed it can be proven — that most of them are nonsense. Where gradients exist, they are unrelated to real solutions. On thes... |

12 |
Knowledge-Based Self-Adaptation in Evolutionary Programming Using Cultural Algorithms
- Reynolds, Chung
- 1997
(Show Context)
Citation Context ...f co-sinuses, making it much harder to find the global minimum), Rosenbrock’s function in 30 dimensions, and Schaffer’s f6, which is in 2 dimensions. Formulas can be found in the literature, e.g., in =-=[5]-=-. It does not seem interesting to us to demonstrate that an algorithm is good on some functions and not on others. What we hope for is a problem-solver that can work well with a wide range of problems... |

8 |
EPSO-Best-of-Two-Worlds Meta-Heuristic Applied to Power System Problems
- Miranda, Fonseca
- 2002
(Show Context)
Citation Context ...vidual’s own previous best needs to be part of the formula. But it must be noted that the particle’s own experience does contribute information to its trajectory, in the form that Miranda and Fonseca =-=[13]-=- call its “habit.” For instance, the current velocity is created out of the velocities of the previous time steps as the particle maintains a cyclic trip through the search space. The particle’s curre... |

3 |
Free Lunch Theorems for Optimization
- “No
- 1997
(Show Context)
Citation Context ... functions and not on others. What we hope for is a problem-solver that can work well with a wide range of problems. This line of reasoning drives us head-on into the No Free Lunch theorem (NFL) [6], =-=[7]-=-. A. Free Lunch NFL asserts that no algorithm can be better than any other, over all possible functions. This seems to be true because of two classes of functions: deceptive ones, and random ones. Dec... |