K-Ortalamalara dayalı yeni yoğunlaştırma stratejisi ile Parçacık Sürüsü Optimizasyonu

Parçacık Sürü Optimizasyonu (PSO), sürü zekâsı temelli metaheuristik algoritmadır. PSO, balıkların veya kuşların yiyecek arama davranışlarından esinlenilerek modellenmiştir. PSO, sade ve etkili bir çalışma yapısına sahip olmasının avantajlarına rağmen, erken yakınsama, yerel minimuma takılma ve zayıf küresel arama kapasitesi gibi bazı dezavantajları da bulunmaktadır. Bu çalışmada, PSO'nun performansını artırmak için K-Ortalamalar kümelemeye dayalı yeni bir yoğunlaştırma stratejisi önerilmiştir. Önerilen yönteme, K-Ortalamalara Dayalı Yeni Yoğunlaştırma Stratejisi ile Parçacık Sürüsü Optimizasyonu (PSO-ISK) adı verilmiştir. PSO-ISK'nın ilk adımında, PSO'daki parçacıklar farklı kümelere ayrılmaktadır. Sonraki adımda ise, her küme için bir merkez ve merkeze en uzak parçacık belirlenmektedir. Bu çalışmanın sonucunda, PSO-ISK, merkeze en uzak parçacığın sonuçlarını iyileştirerek yeni bir yoğunlaştırma stratejisi önermektedir. PSO-ISK'nın performansı, 16 farklı benchmark test fonksiyonu kullanılarak sonuçlar analiz edilmiştir. Elde edilen sonuçlar, Standart PSO (SPSO) ve 7 farklı PSO varyantı ile karşılaştırılmıştır. Yapılan karşılaştırmalar sonucunda, PSO-ISK'nın diğer çalışmalara göre daha etkili sonuçlar elde ettiği ve PSO-ISK'nın performans iyileştirmesindeki önemi kanıtlamıştır.

Anahtar Kelimeler:

K-Ortalamalar, PSO, Yoğunlaştırma stratejisi

Particle Swarm Optimization with a new intensification strategy based on K-Means

Particle Swarm Optimization (PSO) is a swarm intelligence-based metaheuristic algorithm inspired by the foraging behaviors of fish or birds. Despite the advantages of having a simple and effective working structure, PSO also has some disadvantages, such as early convergence, getting trapped in local minima, and weak global search capabilities. In this study, a novel intensification strategy based on K-Means clustering has been proposed to enhance the performance of PSO. The proposed method is called Particle Swarm Optimization with a New Intensification Strategy based on K-Means (PSO-ISK). In the first step of PSO-ISK, particles in PSO are grouped into different clusters. Then, a center and the farthest particle from the center are identified for each cluster. PSO-ISK proposes a new intensification strategy by improving the results of the farthest particle from the center. The performance of PSO-ISK is analyzed using 16 different benchmark test functions. The obtained results are compared with Standard PSO (SPSO) and 7 different PSO variants. According to the comparison results, PSO-ISK provides a notable performance improvement by outperforming SPSO and all seven PSO variants. The comparisons conducted have proven that PSO-ISK produces more effective outcomes than other studies, which results in a significant contribution to improving performance.

Keywords:

Intensification strategy, K-Means, PSO,

PDF

___

[1] MiarNaeimi F, Azizyan G, Rashki M. "Horse herd optimization algorithm: A nature-inspired algorithm for high-dimensional optimization problems". Knowledge-Based Systems, 213, 1-17, 2021.
[2] Chou JS, Truong DN. "A novel metaheuristic optimizer inspired by behavior of jellyfish in ocean". Applied Mathematics and Computation, 389, 1-47, 2021.
[3] Ahmadianfar I, Bozorg-Haddad O, Chu X. "Gradient-based Optimizer: A new metaheuristic optimization algorithm". Information Sciences, 540, 131-159, 2020.
[4] Askari Q, Younas I, Saeed M. "Political optimizer: A novel socio-inspired meta-heuristic for global optimization". Knowledge-Based Systems, 195, 1-25, 2020.
[5] Shadravan S, Naji HR, Bardsiri VK. "The Sailfish Optimizer: A novel nature-inspired metaheuristic algorithm for solving constrained engineering optimization problems". Engineering Applications of Artificial Intelligence, 80, 20-34, 2019.
[6] Arora S, Singh S. "Butterfly optimization algorithm: a novel approach for global optimization". Soft Computing, 23(3), 715-734, 2019.
[7] Kiran MS. "TSA: Tree-seed algorithm for continuous optimization". Expert Systems with Applications, 42(19), 6686-6698, 2015.
[8] James JQ, Li VO. "A social spider algorithm for global optimization". Applied Soft Computing, 30, 614-627, 2015.
[9] Kennedy J, Eberhart R. "Particle swarm optimization". Proceedings of ICNN'95-International Conference on Neural Networks, Perth, WA, Australia, 27 November-01 December 1995.
[10] Haklı H, Uğuz H. "A novel particle swarm optimization algorithm with levy flight". Applied Soft Computing, 23, 333-345, 2014.
[11] Lei L, Min X, Xiaokui L. "Research on hybrid PSO algorithm with appended intensification and diversification". Proceedings 2013 International Conference on Mechatronic Sciences, Electric Engineering and Computer (MEC), Shenyang, China, 20-22 December 2013.
[12] Liang, JJ, Qin AK, Suganthan PN, Baskar S. "Comprehensive learning particle swarm optimizer for global optimization of multimodal functions". IEEE Transactions on Evolutionary Computation, 10(3), 281-295, 2006.
[13] Ratnaweera A, Halgamuge SK, Watson HC. "Selforganizing hierarchical Particle Swarm Optimizer with Time-varying acceleration coefficients". IEEE Transactions on Evolutionary Computation, 8(3), 240-255, 2004.
[14] Mendes R, Kennedy J, Neves J, "The fully informed Particle Swarm: simpler, maybe better". IEEE Transactions on Evolutionary Computation, 8(3), 204-210, 2004.
[15] Kennedy J, Mendes R. "Population structure and particle swarm performance". Proceedings of the 2002 Congress on Evolutionary Computation CEC'02, (Cat. No.02TH8600), Honolulu, HI, USA, 12-17 May 2002.
[16] Liang JJ, Suganthan PN. "Dynamic multi-swarm Particle Swarm Optimizer". Proceedings 2005 IEEE Swarm Intelligence Symposium, Pasadena, CA, USA, 08-10 June 2005.
[17] Solaiman BF, Sheta A. "Energy optimization in wireless sensor networks using a hybrid K-means PSO clustering algorithm". Turkish Journal of Electrical Engineering & Computer Sciences, 24(4), 2679-2695, 2016.
[18] Gao H, Li Y, Kabalyants P, Xu H, Martinez-Bejar R. "A novel hybrid PSO-K-Means clustering algorithm using gaussian estimation of distribution method and Lévy Flight". IEEE Access, 8, 122848-122863, 2020.
[19] Mahajan M, Kumar S, Pant B. "Prediction of environmental pollution using hybrid PSO-K-Means approach". International Journal of E-Health and Medical Communications (IJEHMC), 12(2), 65-76, 2021.
[20] Sun Y, Liu G, Zheng D, Zou H, Liu Z, Liu J. "A self-adaptive Particle Swarm Optimization based K-means (SAPSO-K) clustering method to evaluate fabric tactile comfort". The Journal of the Textile Institute, 4(1), 1-12, 2021.
[21] Younus ZS, Mohamad D, Saba T, Alkawaz MH, Rehman A, Al-Rodhaan M, Al-Dhelaan A. "Content-based image retrieval using PSO and K-means clustering algorithm". Arabian Journal of Geosciences, 8(8), 6211-6224, 2015.
[22] Liu S, Zou Y. "An improved hybrid clustering algorithm based on Particle Swarm Optimization and K-means". IOP Conference Series: Materials Science and Engineering, Singapore, 27-29 February 2020.
[23] Jamali-Dinan SS, Soltanian-Zadeh H, Bowyer SM, Almohri H, Dehghani H, Elisevich K, Nazem-Zadeh MR. "A combination of particle swarm optimization and minkowski weighted K-Means clustering: application in lateralization of temporal lobe epilepsy". Brain Topography, 33(4), 519-532, 2020.
[24] Tarkhaneh O, Isazadeh A, Khamnei HJ. "A new hybrid strategy for data clustering using cuckoo search based on Mantegna Levy distribution, PSO and K-means". International Journal of Computer Applications in Technology, 58(2), 137-149, 2018.
[25] MacQueen, J. “Some methods for classification and analysis of multivariate observations”. Proceedings of the 5th Berkeley Symposium on Mathematical Statistics and Probability, Los Angeles, USA, 1 January 1967.
[26] Particle Swarm Central. ''SPSO 2007 Matlab''. http://www.particleswarm.info/Programs.html (07.07.2020).
[27] Karaboga D, Akay B. "A comparative study of artificial bee colony algorithm". Applied mathematics and computation, 214(1), 108-132, 2009.
[28] Zhan Z, Zhang J, Li Y, Shi Y. "Orthogonal learning particle swarm optimization". IEEE Transactions on Evolutionary Computation, 15(6), 832-847, 2010.