Particle swarm optimization with combined mutation and hill climbing
نویسنده
چکیده
Particle swarm optimization is an established, effective technique for global optimization, but one which is prone to delivering sub-optimal results when faced with multimodal functions, especially when the local optima are found in regions where the objective function varies rapidly. The failure of the swarm to venture into these regions is not just of theoretical interest; it has been observed in practical settings, e.g. in the training of neural networks. In this paper we propose a simple modification to the particle swarm algorithm in which, at each generation, a small number of particles are mutated and are allowed to hill climb. The mutation has the effect of randomly bouncing particles towards other parts of the search space, while the effect of hill-climbing is to greatly increase the effective size of the “target” region of interest around the global optimum. We provide results for a selection of well-known test functions, and demonstrate that our modification improves the ability of the swarm to find the global optimum. In most cases this is achieved without a prohibitive increase in the number of function evaluations.
منابع مشابه
Multi-phase Discrete Particle Swarm Optimization
This paper describes a successful adaptation of the Particle Swarm Optimization algorithm to discrete optimization problems. In the proposed algorithm, particles cycle through multiple phases with differing goals. We also exploit hill climbing. On benchmark problems, this algorithm outperforms a genetic algorithm and a previous discrete PSO formulation.
متن کاملImproving Particle Swarm Optimization by hybridization of stochastic search heuristics and Self-Organized Criticality
The objective of this thesis is to investigate how to improve Particle Swarm Optimization by hybridization of stochastic search heuristics and by a Self-Organized Criticality extension. The thesis will describe two hybrid models extending Particle Swarm Optimization with two aspects from Evolutionary Algorithms, recombination via breeding and gene flow restriction via subpopulations. A further ...
متن کاملUsing Artificial Intelligence Techniques for Strategy Generation in the Commons Game
In this paper, we consider the use of artificial intelligence techniques to aid in discovery of winning strategies for the Commons Game (CG). The game represents a common scenario in which multiple parties share the use of a self-replenishing resource. The resource deteriorates quickly if used indiscriminately. If used responsibly, however, the resource thrives. We consider the scenario one pla...
متن کاملThe LifeCycle Model: Combining Particle Swarm Optimisation, Genetic Algorithms and HillClimbers
Adaptive search heuristics are known to be valuable in approximating solutions to hard search problems. However, these techniques are problem dependent. Inspired by the idea of life cycle stages found in nature, we introduce a hybrid approach called the LifeCycle model that simultaneously applies genetic algorithms (GAs), particle swarm optimisation (PSOs), and stochastic hill climbing to creat...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2004