Experimental Results for the study "A Modular Hybridization of Particle Swarm Optimization and Differential Evolution"

This repository contains the experiment results and R scripts to analyze the data for the study "A Modular Hybridization of Particle Swarm Optimization andDifferential Evolution", which is accepted in The Genetic and Evolutionary Computation Conference (GECCO) '20 conference: 

Rick Boks, Hao Wang, and Thomas Bäck. 2020. A Modular Hybridization of Particle Swarm Optimization and Differential Evolution. In Genetic and Evolutionary Computation Conference Companion (GECCO ’20 Companion), July 8–12, 2020, Cancún, Mexico. ACM, New York, NY, USA, 8 pages. https: //doi.org/10.1145/3377929.3398123

Bibtex:

@inproceedings{BoksWB20,
author = {Rick Boks and Hao Wang and Thomas B\"ack},
title = {{A Modular Hybridization of Particle Swarm Optimization and
Differential Evolution}},
booktitle = {Proceedings of the Genetic and Evolutionary Computation Conference,
{GECCO} 2020, Canc\'un, Mexico, July 8-12, 2020},
publisher = {{ACM}},
year = {2020},
url = {https://doi.org/10.1145/3321707.3321816},
doi = {doi.org/10.1145/3377929.3398123,
}

Data description: we benchmarked 800 different hybridizations of the Particle Swarm Optimization (PSO) and Differential Evolution (DE) algorithms on a well-known continuous black-box problem set called COCO/BBOB, which consists of 24 test functions. 30 independent runs are conducted for each algorithm on each problem.

'ERT.csv': a data frame with columns DIM (5D or 20D), funcId (F1-24), algId (algorithm names), target (\(10^{\{-8,-7, \ldots, 1\}}\)), ERT (expected running time), and sd (standard deviation).

'raw-data.csv': the running time recorded in each independent run. 

'analysis.R': the R script that generates ERT tables in the paper.

'ecdf.R': the R script that renders the ECDF (empirical cumulative distribution function) plots in the paper.

本仓库包含为研究《粒子群优化与差分进化的模块化混合策略》开展的实验结果与数据分析R脚本，该研究已被2020年遗传与进化计算会议（Genetic and Evolutionary Computation Conference, GECCO '20）收录：

Rick Boks、王浩与Thomas Bäck。2020年。粒子群优化与差分进化的模块化混合策略。见：遗传与进化计算会议Companion（GECCO ’20 Companion），2020年7月8日至12日，墨西哥坎昆。ACM出版社，美国纽约州纽约市，共8页。https://doi.org/10.1145/3377929.3398123

Bibtex引用：
@inproceedings{BoksWB20,
author = {Rick Boks and Hao Wang and Thomas Bäck},
title = {{A Modular Hybridization of Particle Swarm Optimization and Differential Evolution}},
booktitle = {Proceedings of the Genetic and Evolutionary Computation Conference,
{GECCO} 2020, Cancún, Mexico, July 8-12, 2020},
publisher = {{ACM}},
year = {2020},
url = {https://doi.org/10.1145/3321707.3321816},
doi = {doi.org/10.1145/3377929.3398123},
}

数据集说明：我们在知名的连续黑箱问题集COCO/BBOB上对粒子群优化（Particle Swarm Optimization, PSO）与差分进化（Differential Evolution, DE）算法的800种不同混合策略进行了基准测试，该测试集包含24个测试函数。每种算法在每个测试问题上均执行30次独立运行。

'ERT.csv'：包含以下列的数据框：DIM（5维或20维）、funcId（F1-24）、algId（算法名称）、target（(10^{{-8,-7, ldots, 1}})）、ERT（期望运行时间）与sd（标准差）。
'raw-data.csv'：每次独立运行记录的运行时间。
'analysis.R'：用于生成论文中ERT表格的R脚本。
'ecdf.R'：用于绘制论文中经验累积分布函数（Empirical Cumulative Distribution Function, ECDF）的绘图脚本。