Prof.dr. Peter A.N. Bosman
Senior Researcher at Centrum Wiskunde & Informatica (CWI)
(Dutch National Research Institute for Mathematics and Computer Science)
Life Sciences and Health Research Group

Professor of Evolutionary Algorithms at Delft University of Technology
Algorithmics Section of the Department of Software Technology
Faculty of Electrical Engineering, Mathematics and Computer Science

E Peter.Bosman@cwi.nl    W http://homepages.cwi.nl/~bosman    T +31(0)20 592 4265
Home
Resume
Publications
Source Code


CWI
DISCLAIMER
Source Code

PermutationGOMEA                                                                                                                                                                 
Description    Source code for the first Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) instance dedicated to permutation optimization. The publication most closely related with this code is:

P.A.N. Bosman, N.H. Luong, and D. Thierens. Expanding from Discrete Cartesian to Permutation Gene-pool Optimal Mixing Evolutionary Algorithms. In T. Friedrich et al., editors, Proceedings of the Genetic and Evolutionary Computation Conference - GECCO-2016, pages 637-644, ACM Press, New York, New York, 2016.
[Nominated for Best Paper Award]


The zip file contains 1 version of Permutation GOMEA, written in C.

Download    PermutationGOMEA.zip   [ Current number of downloads: 129 ]


(MO-)RV-GOMEA                                                                                                                                                                 
Description    Source code for the first Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) instance dedicated to Real-Valued optimization (RV-GOMEA), both single-objective and multi-objective. The publications most closely related with this code are:

A. Bouter, T. Alderliesten, C. Witteveen, and P.A.N. Bosman. Exploiting linkage information in real-valued optimization with the real-valued gene-pool optimal mixing evolutionary algorithm. In G. Ochoa et al., Proceedings of the Genetic and Evolutionary Computation Conference - GECCO-2017, pages 705-712, ACM Press, New York, New York, 2017.

A. Bouter, N.H. Luong, C. Witteveen, T. Alderliesten, and P.A.N. Bosman. The multi-objective real-valued gene-pool optimal mixing evolutionary algorithm. In G. Ochoa et al., Proceedings of the Genetic and Evolutionary Computation Conference - GECCO-2017, pages 537-544, ACM Press, New York, New York, 2017.

The zip file contains 2 versions of RV-GOMEA (single-objective and multi-objective), written in C.

Download    RV-GOMEA.zip   [ Current number of downloads: 270 ]


GP-GOMEA                                                                                                                                                                             
Description    Source code for the first Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) instance dedicated to Genetic Programming (GP-GOMEA). The publication most closely related with this code is:

M. Virgolin, T. Alderliesten, C. Witteveen, and P.A.N. Bosman. Scalable genetic programming by gene-pool optimal mixing and input-space entropy-based building block learning. In G. Ochoa et al., Proceedings of the Genetic and Evolutionary Computation Conference - GECCO-2017, pages 1041-1048, ACM Press, New York, New York, 2017.

The zip file contains 1 version of GP-GOMEA, written in C++.

Download    GP-GOMEA.zip   [ Current number of downloads: 232 ]


MO-GOMEA                                                                                                                                                                             
Description    Source code for the Binary Multi-Objective Gene-pool Optimal Mixing Evolutionary Algorithm (MO-GOMEA). The publication most closely related with this code is:

N.H. Luong, H. La Poutré, and P.A.N. Bosman. Multi-objective Gene-pool Optimal Mixing Evolutionary Algorithm with the Interleaved Multi-start Scheme. In Swarm and Evolutionary Computation 40, pages 238-254, 2018.

The zip file contains 1 version of MO-GOMEA, written in C++.

Download    MO-GOMEA.zip   [ Current number of downloads: 66 ]


LTGA                                                                                                                                                                                       
Description    NOTE: THIS IS NOT THE RECENTMOST VERSION OF BINARY GOMEA. Code for that version is best obtained by taking PermutationGOMEA and changing the random keys representation into the binary representation as found in LTGA.zip. Sorry for the inconvenience. We are working on a more complete GOMEA platform code that should remain up-to-date. Source code for the Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) instance that is also known as the Linkage Tree Genetic Algorithm (LTGA). The publication most closely related with this code is:

P.A.N. Bosman and D. Thierens. More Concise and Robust Linkage Learning by Filtering and Combining Linkage Hierarchies. In C. Blum and E. Alba, editors, Proceedings of the Genetic and Evolutionary Computation Conference - GECCO-2013, pages 359-366, ACM Press, New York, New York, 2013.

The zip file contains 1 version of LTGA, written in C.

Download    LTGA.zip   [ Current number of downloads: 656 ]


AMaLGaM                                                                                                                                                                               
Description    Source code for the Estimation-of-Distribution Algorithm (EDA) known as the Adapted Maximum-Likelihood Gaussian Model Iterated Density Estimation Evolutionary Algorithm (AMaLGaM-IDEA, or AMaLGaM for short). The publication most closely related with this code is:

P.A.N. Bosman, J. Grahl and D. Thierens. Benchmarking Parameter-free AMaLGaM on Functions With and Without Noise . In Evolutionary Computation 21(3), pages 445-469, 2013.

The zip file contains 12 versions of AMaLGaM. The Gaussian distribution is either not factorized, meaning a full covariance matrix is used, factorized using a Bayesian factorization that is learned in a greedy fashion, or factorized using the univariate factorization. For each of these three factorizations, the distribution can either be learned incrementally or not. Finally, there is a parameter-free version of each EDA with an automated restart mechanism. The source code is written in C.

Download    AMaLGaM.zip   [ Current number of downloads: 1260 ]