Multi-objective particle swarm optimisation methods
نویسنده
چکیده
This study compares a number of selection regimes for the choosing of global best (gbest) and personal best (pbest) for swarm members in multi-objective particle swarm optimisation (MOPSO). Two distinct gbest selection techniques are shown to exist in the literature, those that do not restrict the selection of archive members and those with ‘distance’ based gbest selection techniques. Theoretical justification for both of these approaches is discussed, in terms of the two types of search that these methods promote, and the potential problem of particle clumping in MOPSO is described. The popular pbest selection methods in the literature are also compared, and the affect of the recently introduced turbulence term is viewed in terms of the additional search it promotes, across all parameter combinations. In light of the discussion, new avenues of MOPSO research are highlighted.
منابع مشابه
Niching for Ant Colony Optimisation
Evolutionary Computation niching methods, such as Fitness Sharing and Crowding, are aimed at simultaneously locating and maintaining multiple optima to increase search robustness, typically in multi-modal function optimization. Such methods have been shown to be useful for both single and multiple objective optimisation problems. Niching methods have been adapted in recent years for other optim...
متن کاملA Particle Swarm Optimisation Based Multi-objective Filter Approach to Feature Selection for Classification
Feature selection (FS) has two main objectives of minimising the number of features and maximising the classification performance. Based on binary particle swarm optimisation (BPSO), we develop a multi-objective FS framework for classification, which is NSBPSO based on multi-objective BPSO using the idea of non-dominated sorting. Two multi-objective FS algorithms are then developed by applying ...
متن کاملA Web-based System for Visualisation-driven Interactive Multi-objective Optimisation
Interactive Multi-Objective Optimisation is an increasing field of evolutionary and swarm intelligence-based algorithms. By involving a human decision, a set of relevant non-dominated points can often be acquired at significantly lower computational costs than with a posteriori algorithms. A rarely addressed issue in interactive optimisation is the design of efficient user interfaces and the ap...
متن کاملSimilarity based Multi-objective Particle Swarm Optimisation for Feature Selection in Classification
This paper presents a particle swarm optimisation (PSO) based multi-objective feature selection approach to evolving a set of non-dominated feature subsets and achieving high classification performance. Firstly, a pure multi-objective PSO (named MOPSO-SRD) algorithm, is applied to solve feature selection problems. The results of this algorithm is then used to compare with the proposed a multi-o...
متن کاملAsynchronous Multi-Objective Optimisation in Unreliable Distributed Environments
This chapter examines the performance characteristics of both asynchronous and synchronous parallel particle swarm optimisation algorithms in heterogeneous, fault-prone environments. The chapter starts with a simple parallelisation paradigm, the Master-Slave model using Multi-Objective Particle Swarm Optimisation (MOPSO) in a heterogeneous environment. Extending the investigation to general, di...
متن کاملClustering-Based Leaders' Selection in Multi-Objective Particle Swarm Optimisation
Clustering-based Leaders’ Selection (CLS) is a novel approach for leaders selection in multi-objective particle swarm optimisation. Both objective and solution spaces are clustered. An indirect mapping between clusters in both spaces is defined to recognize regions with potentially better solutions. A leaders archive is built which contains representative particles of selected clusters in the o...
متن کامل