Inertia weight control strategies for particle swarm optimization

Harrison, Kyle Robert; Engelbrecht, Andries P.; Ombuki-Berman, Beatrice M.

doi:10.1007/s11721-016-0128-z

Cited by 84 publications

(47 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The non-linear relationship between the movement patterns of particles and the coefficients, found in this article, indicate that the assumptions used for proposing adaptive approaches in PSO (e.g., [28], [25]) were somewhat simplistic. This provides theoretical justification for the findings in [9] where it was experimentally (on a rather large set of benchmark functions) shown that none of the PSO-based adaptive approaches tested in that study can beat a PSO with constant coefficients, that is indeed counterintuitive.…”

Section: Discussionmentioning

confidence: 55%

“…In this section we compare 14 algorithms, 2 with constant coefficients, 12 with adaptive or time-adaptive coefficients, against a time adaptive approach based on analyses conducted in this paper. The methods for comparison are • Top 9 methods in [9] that are: Constriction coefficient PSO (CCPSO) [10], Linear decreasing inertia weight PSO (LDWPSO) [28], Random inertia weight PSO (RWPSO) [31], Chaotic descending inertia weight PSO [13], sugeno inertia weight PSO [32], logarithm decreasing PSO [33], self-regulating PSO 9 The aim of this article is not to find the best values for the coefficients to design yet another adaptive PSO to beat other existing PSO. The main aim of this article is to theoretically support considerations that need to be factored in for any adaptive PSO to be designed.…”

Section: Experiments and Comparisonsmentioning

confidence: 99%

See 1 more Smart Citation

A Theoretical Guideline for Designing an Effective Adaptive Particle Swarm

Bonyadi

2020

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

In this paper we theoretically investigate underlying assumptions that have been used for designing adaptive particle swarm optimization algorithms in the past years. We relate these assumptions to the movement patterns of particles controlled by coefficient values (inertia weight and acceleration coefficient) and introduce three factors, namely the autocorrelation of the particle positions, the average movement distance of the particle in each iteration, and the focus of the search, that describe these movement patterns. We show how these factors represent movement patterns of a particle within a swarm and how they are affected by particle coefficients (i.e., inertia weight and acceleration coefficients). We derive equations that provide exact coefficient values to guarantee achieving a desired movement pattern defined by these three factors within a swarm. We then relate these movements to the searching capability of particles and provide guideline for designing potentially successful adaptive methods to control coefficients in particle swarm. Finally, we propose a new simple time adaptive particle swarm and compare its results with previous adaptive particle swarm approaches. Our experiments show that the theoretical findings indeed provide a beneficial guideline for successful adaptation of the coefficients in the particle swarm optimization algorithm.

show abstract

Section: Discussionmentioning

confidence: 55%

Section: Experiments and Comparisonsmentioning

confidence: 99%

A Theoretical Guideline for Designing an Effective Adaptive Particle Swarm

Bonyadi

2020

IEEE Trans. Evol. Computat.

View full text Add to dashboard Cite

show abstract

“…where k indicates the vector component, (α t ), (β t ), and (γ t ) are sequences of random variables. The class of PSOs described by equation (49) includes numerous PSO variants where the inertia, cognitive and/or social coefficients are altered over time, as in many self-adaptive PSOs (Naka et al, 2001;Ratnaweera et al, 2003;Suganthan, 1999;Yoshida et al, 1999;Perman et al, 2003;Harrison et al, 2016). The second relatively unexplored area is to perform theoretical stability analysis on PSO variants where the particle position update equation does not operate on dimensions independently.…”

Section: Discussionmentioning

confidence: 99%

Particle swarm stability: a theoretical extension using the non-stagnate distribution assumption

Cleghorn

Engelbrecht

2017

Swarm Intell

Self Cite

View full text Add to dashboard Cite

This paper presents an extension of the state of the art theoretical model utilized for understanding the stability criteria of the particles in particle swarm optimization algorithms. Conditions for order-1 and order-2 stability are derived by modeling, in the simplest case, the expected value and variance of a particle's personal and neighborhood best positions as convergent sequences of random variables. Furthermore, the condition that the expected value and variance of a particle's personal and neighborhood best positions are convergent sequences is shown to be a necessary condition for order-1 and order-2 stability. The theoretical analysis presented is applicable to a large class of particle swarm optimization variants.

show abstract

“…Mohammadi et al [10] applied the SVM optimized by the PSO (PSO-SVM) to detect the squamous disease, with a recognition accuracy of 98.9%. Due to the weak local optimization ability of the conventional PSO algorithm, the inertia weight factor has been proposed to improve the optimization ability of the PSO algorithm [11].…”

Section: Introductionmentioning

confidence: 99%

An Improved Skewness Decision Tree SVM Algorithm for the Classification of Steel Cord Conveyor Belt Defects

Mao

Zhang

et al. 2018

Applied Sciences

View full text Add to dashboard Cite

Skewness Decision Tree Support Vector Machine (SDTSVM) algorithm is widely known as a supervised learning model for multi-class classification problems. However, the classification accuracy of the SDTSVM algorithm depends on the perfect selection of its parameters and the classification order. Therefore, an improved SDTSVM (ISDTSVM) algorithm is proposed in order to improve the classification accuracy of steel cord conveyor belt defects. In the proposed model, the classification order is determined by the sum of the Euclidean distances between multi-class sample centers and the parameters are optimized by the inertia weight Particle Swarm Optimization (PSO) algorithm. In order to verify the effectiveness of the ISDTSVM algorithm with different feature space, experiments were conducted on multiple UCI (University of California Irvine) data sets and steel cord conveyor belt defects using the proposed ISDTSVM algorithm and the conventional SDTSVM algorithm respectively. The average classification accuracies of five-fold cross-validation were obtained, based on two kinds of kernel functions respectively. For the Vowel, Zoo, and Wine data sets of the UCI data sets, as well as the steel cord conveyor belt defects, the ISDTSVM algorithm improved the classification accuracy by 3%, 3%, 1% and 4% respectively, compared to the SDTSVM algorithm. The classification accuracy of the radial basis function kernel were higher than the polynomial kernel. The results indicated that the proposed ISDTSVM algorithm improved the classification accuracy significantly, compared to the conventional SDTSVM algorithm.

show abstract

Inertia weight control strategies for particle swarm optimization

Cited by 84 publications

References 50 publications

A Theoretical Guideline for Designing an Effective Adaptive Particle Swarm

A Theoretical Guideline for Designing an Effective Adaptive Particle Swarm

Particle swarm stability: a theoretical extension using the non-stagnate distribution assumption

An Improved Skewness Decision Tree SVM Algorithm for the Classification of Steel Cord Conveyor Belt Defects

Contact Info

Product

Resources

About