As computing power grows, the automated specialization and design of evolutionary algorithms (EAs) to tune their performance to individual problem classes becomes more attractive. To this end, a significant amount of ...
详细信息
ISBN:
(纸本)9781450383509
As computing power grows, the automated specialization and design of evolutionary algorithms (EAs) to tune their performance to individual problem classes becomes more attractive. To this end, a significant amount of research has been conducted in recent decades, utilizing a wide range of techniques. However, few techniques have been devised which automate the design of the overall structure of an EA. Most EA implementations rely solely on the traditional evolutionary cycle of parent selection, reproduction, and survival selection, and those with unique structures typically replace this with another static, hand-made cycle. Existing techniques for modifying the evolutionary structure use representations which are either loosely structured and highly stochastic, or which are constrained and unable to easily evolve complicated pathways. The ability to easily evolve complex evolutionary pathways would greatly expand the heuristic space which can be explored during specialization, potentially allowing for the representation of EAs which outperform the traditional cycle. This work proposes a methodology for the automateddesign of the evolutionary process by introducing a novel directed-graph-based representation, created to be mutable and flexible, permitting a black-box designer to produce reusable, high-performance EAs. Experiments show that our methodology can produce high-performance EAs demonstrating intelligible strategies.
Metaheuristics are an effective and diverse class of optimization algorithms: a means of obtaining solutions of acceptable quality for otherwise intractable problems. The selection, construction, and configuration of ...
详细信息
Metaheuristics are an effective and diverse class of optimization algorithms: a means of obtaining solutions of acceptable quality for otherwise intractable problems. The selection, construction, and configuration of a metaheuristic for a given problem has historically been a manually intensive process based on experience, experimentation, and reasoning by metaphor. More recently, there has been interest in automating the process of algorithm configuration. In this article, we identify shared state as an inhibitor of progress for such automation. To solve this problem, we introduce the automated Open-Closed Principle (AOCP), which stipulates design requirements for unintrusive reuse of algorithm frameworks and automated assembly of algorithms from an extensible palette of components. We demonstrate how the AOCP enables a greater degree of automation than previously possible via an example implementation.
This article proposes a simplistic algorithmic framework, namely hyperSPAM, composed of three search algorithms for addressing continuous optimisation problems. The Covariance Matrix Adaptation Evolution Strategy (CMA...
详细信息
This article proposes a simplistic algorithmic framework, namely hyperSPAM, composed of three search algorithms for addressing continuous optimisation problems. The Covariance Matrix Adaptation Evolution Strategy (CMAES) is activated at the beginning of the optimisation process as a preprocessing component for a limited budget. Subsequently, the produced solution is fed to the other two single-solution search algorithms. The first performs moves along the axes while the second makes use of a matrix orthogonalization to perform diagonal moves. Four coordination strategies, in the fashion of hyperheuristics, have been used to coordinate the two single-solution algorithms. One of them is a simple randomized criterion while the other three are based on a success based reward mechanism. The four implementations of the hyperSPAM framework have been tested and compared against each other and modern metaheuristics on an extensive set of problems including theoretical functions and real-world engineering problems. Numerical results show that the different versions of the framework display broadly a similar performance. One of the reward schemes appears to be marginally better than the others. The simplistic random coordination also displays a very good performance. All the implementations of hyperSPAM significantly outperform the other algorithms used for comparison. (C) 2018 Elsevier Inc. All rights reserved.
In the last 20 years, literally dozens of optimization algorithms based on swarm intelligence have been proposed. Particle Swarm Optimization, Artificial Bee Colony, Cuckoo Search, Firefly Optimization, and Cat Swarm ...
详细信息
ISBN:
(纸本)9781450367486
In the last 20 years, literally dozens of optimization algorithms based on swarm intelligence have been proposed. Particle Swarm Optimization, Artificial Bee Colony, Cuckoo Search, Firefly Optimization, and Cat Swarm Optimization are just a small sample of the exuberance of swarm-like algorithms. Although they differ in implementation details, they all share a common structure: an update rule is applied to each solution, followed by a drop rule that decides whether to keep the updated solution or not. In this poster we explore the idea of automatically generating swarm-like optimizers. Our proposal is divided in two stages: First we decompose popular, human-crafted, swarm-like optimizers such as PSO, CS, ABC (as well as DE/GA) into a list of basic rules. Second, we use Grammatical Evolution to procedurally generate variations on this base structure by recombining these operators. We generate three instances of algorithms, and observe that they have comparable performance to DE and PSO. Our framework will be useful to gain insight on the design space of meta-heuristics and the nature of swarm-like algorithms.
暂无评论