READMe File for Dataset for the thesis for the degree of Doctorate of Philosophy, "Improving the applicability of genetic algorithms to real problems", by Przemyslaw Andrzej Grudniewski.

Dataset DOI is 10.5258/SOTON/D1709

This dataset supports the thesis entitled "Improving the applicability of genetic algorithms to real problems"
AWARDED BY: Univeristy of Southampton
DATE OF AWARD: 2021

- The Dataset is separated according to chapters of the Thesis
- In each chapter the data is separated into 3 main categories: a) Figures in folder "Figures"; b) Table data in folder "Processed Table Data"; and c) raw data files in folder "Raw Data"

ad. a) Figures:
- all figures used in thesis are in the main folder. Figures containing Pareto Optimal Fronts are named "Algorithm name + function name + PF.jpg". Figures with contour plots are named "Algorithm name + function name + Contour_Plot.jpg".
- Folder "Data" contains raw (float) data of Pareto Optimal Fronts obtained by algorithms (names are corresponding to the figures).
- Folder "Additional Figures" contains figures for additional cases (functions/algorithms) not included in the main part of the thesis.

ad. b) Tables:
- main .xlsx file contain processed data for tables used in thesis, as well as the data for similar cases (different functions/algorithms), not included in the main part of the thesis.
- Folder "Input" contains raw data, used to created the given .xlsx file. The explanation of raw data is provided in ad. c)

ad. c) Raw data:
- This folder contains all data generated for each GA simulation used for the purposes of the thesis.
- The folders are named according to following rule: "date + time + name of the simulation + GA parameters". GA parameters are population size, number of iterations, number of collectives, frequency of collectives elimination.
- Inside the folders the data is provided for each run as: graphical representation of each point found on the objective space as "#run number + Graph_all.jpg"; graphical representation of found Pareto optimal front as "#run number + PF.jpg"; the float data for each point of Pareto optimal front in "#run number + PF.csv" - in the first column contains the values of every objective, other columns contain values of each variable.
- File "folder name + .xslx" contains the processed data for the given series of simulations where: Tab "INDEX" contains the parameters setting for each run; Tab "Time_IGD" contains the calculation time and calculated IGD/HV values for each run; Tab "GA_data" contains summarised data for each separate function and algorithm (as all simulations are performed over series of runs in order to find average, min, max and std IGD/HV values).

Date of data collection: 2015.11.02-2020.01.01

Licence:
The dataset is distibuted under Attribution 4.0 International (CC BY 4.0) lincence (https://creativecommons.org/licenses/by/4.0/legalcode).
The attached code within "Additional Data + Code.7z" archive, is distributed under GNU GENERAL PUBLIC LICENSE Version 3, 29 June 2007. Full copyright lincence is included in the archive.

Related projects/Funders:
This research was supported by Lloyd Register Foundation.

Related publications:
P.A. Grudniewski* and A. J. Sobey, “Multi-Level Selection Genetic Algorithm applied to CEC\textquotesingle 09 test instances", 2017 IEEE Congress on Evolutionary Computation (CEC), San Sebastian, pp. 1613-1620, 2017.		
A. J. Sobey* and P.A. Grudniewski, “Re-inspiring the genetic algorithm with multi-level selection theory: multi-level selection genetic algorithm (MLSGA)", Bioinspiration and Biomimetics, vol. 13, no. 5, pp. 1-13, 2018.
P. A. Grudniewski* and A. J. Sobey, “Behaviour of Multi-Level Selection Genetic Algorithm (MLSGA) using different individual-level selection mechanisms”, Swarm and Evolutionary Computation, vol. 44, no. September 2018, pp. 852–862, 2018.
P. A. Grudniewski* and A. J. Sobey “Do general Genetic Algorithms provide benefits when solving real problems?", 2019 IEEE Congress on Evolutionary Computation (CEC), Wellington, 2019.
P. A. Grudniewski* and A. J. Sobey, “cMLSGA: co-evolutionary Multi-Level Selection Genetic Algorithm”, IEEE Transactions on Evolutionary Computation, 2021, Under review.

Repositories on Github (https://github.com/pag1c18/cMLSGA) and Bitbucket (https://bitbucket.org/Pag1c18/)

Date that the file was created: 01.2021