 |
|
 |
|
Software
- XCSF-Ellipsoids Java plus Visualization
Author(s): Patrick O. Stalph and Martin V. Butz (2008)
Language: Java
Description: XCSF-Ellipsoids Java is an XCSF learning classifier system implementation using hyperellipsoidal conditions and recursive least squares predictions for function approximation. The code can be used to evaluate XCSF on several test functions with online visualization support for performance, prediction, and conditions. Other test functions or approximation problems can be easily implemented. See MEDAL Report No. 2008008 for more information.
Download
Documentation - NK Landscapes: Generator of random instances, branch-and-bound solver, and genetic algorithm
Author(s): Martin Pelikan (2008)
Language: ANSI C
Description: This package includes three main parts: (1) A generator for random problem instances of the NK landscape model, (2) a branch-and-bound complete algorithm for NK landscapes, and (3) genetic algorithm code illustrating the use of the code in one's own optimization method (the genetic algorithm is provided only as an example). See MEDAL Report No. 2008001 for more information about the generator or the branch and bound algorithm. Documentation is provided in the package.
Some instances generated with this generator and solved with the provided branch-and-bound solver can be downloaded here: nk-instances.tar.gz. Several tens of thousands instances are provided.
Download
Documentation - XCSFJava
Author(s): Martin V. Butz (2007)
Version: 1.1
Language: Java
Description: Java implementation of the XCS learning classifier system for function approximation (XCSF). The implementation requires requires Java3D (https://java3d.dev.java.net/binary-builds.html).
Download
Documentation - Dependency-Tree Estimation of Distribution Algorithm (dtEDA)
Author(s): Martin Pelikan (2006)
Version: 1.1
Language: C/C++
Description: Implementation of the dependency-tree estimation of distribution algorithm in C/C++. The implementation can deal with alphabets of arbitrary cardinality. If you use version 1.0, you should upgrade to the current version.
Download
Documentation - Generator of Random Additively Decomposable Problems
Author(s): Martin Pelikan, Kumara Sastry, Martin V. Butz, and David E. Goldberg (2006)
Version: 1.0
Language: ANSI C
Description: This package includes the source code of the generator of random additively decomposable problems and additional functions necessary for using the generated problem instances in your own code. An example solver using simple hill climbing is included. Doxygen documentation in html is part of the package.
Download
Documentation - Hierarchical BOA (hBOA)
Author(s): Martin Pelikan and David E. Goldberg (2002)
Language: C/C++
Description: The hierarchical Bayesian optimization algorithm (hBOA) combines BOA, Bayesian networks with local structures, and niching to provide robust and scalable solutions for nearly decomposable and hierarchical problems. A limited demo version is available for download here. Free academic research licences can be provided upon request. For commercial licenses, contact the authors for further information.
Download
Documentation - Bayesian Optimization Algorithm (BOA) with Decision Graphs
Author(s): Martin Pelikan (2000)
Version: 1.1
Language: C/C++
Description: C++ Implementation of the BOA with Decision Graphs.
Download
Documentation - Bayesian Optimization Algorithm (BOA)
Author(s): Martin Pelikan (1999)
Version: 1.0
Language: C/C++
Description: Implementation of the standard BOA with BDe metric in C++.
Download
Documentation
|
Missouri Estimation of Distribution Algorithms Laboratory |
  
|
Web design by Martin Pelikan |
|---|