Projects

Evolving Objects

EO is a template-based, ANSI-C++ evolutionary computation library which helps you to write your own stochastic optimization algorithms insanely fast. With the help of EO, you can easily design evolutionary algorithms that will find solutions to virtually all kind of hard optimization problems ... [More] , from continuous to combinatorial ones. Designing an algorithm with EO consists in choosing what components you want to use for your specific needs, just as building a structure with Lego blocks. [Less]

dANN

The dANN project is a library to help facilitate artificial neural networks within other applications. It is currently written in Java, C++, and C#. However only the java version is currently in active development. If you want to obtain a version other than the java version you will need to get it ... [More] directly from GIT. Our intentions are two fold. First, to provide a powerful interface for programs to include conventional artificial neural network technology into their code. Second, To act as a testing ground for research and development of new AI concepts. We provide new AI technology we have developed, and the latest algorithms already on the market. In the spirit of modular programming the library also provides access to the primitive components giving you greater control. [Less]

GAUL

The Genetic Algorithm Utility Library (or, GAUL for short) is a flexible programming library designed to aid in the development of applications that use genetic, or evolutionary, algorithms. It provides data structures and functions for handling and manipulation of the data required for serial and ... [More] parallel evolutionary algorithms. Additional stochastic algorithms are provided for comparison to the genetic algorithms. [Less]

METSlib Optimization Framework

METSlib is an OO (Object Oriented) metaheuristics framework in C++. Model and algorithms are modular: all the implemented search algorithms can be applied to the same model and personalized algorithms can be applied to very different models. METSlib implements the basics of some metaheuristics ... [More] algorithm: Random Restart Local Search, Variable Neighborhood Search, Iterated Local Search, Simulated Annealing (with linear, exponential and custom cooling schedule), and last but not least Tabu Search. [Less]

MOEA Framework

The MOEA Framework is an open source Java library for developing and experimenting with multiobjective evolutionary algorithms (MOEAs) and other general-purpose optimization algorithms and metaheuristics. A number of algorithms are provided out-of-the-box, including NSGA-II, ε-MOEA, GDE3 and ... [More] MOEA/D. In addition, third-party tools like JMetal and PISA directly integrate with the MOEA Framework. The MOEA Framework targets an academic audience, providing the resources necessary to rapidly design, develop, execute and statistically test optimization algorithms. This includes over 40 test problems from the literature, and a suite of statistical tools for comparing and analyzing algorithm performance. [Less]

evolutionary-algorithm

The package serves as a foundation class library, supporting the implementation many variants of Evolutionary Algorithms, currently including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Memetic Algorithm (MA), Evolution Strategy (ES). Highlighted features · Support both binary ... [More] & real-coded string representations of solution · Operator-based design for flexibility · EA Operators: Selection, Crossover, Mutation, Move operators in PSO & and the adaptive scheme in EA · Individual learning: Davidon–Fletcher–Powell (DFP) and Davies, Swann, and Campey with Gram-Schmidt orthogonalization (DSCG) strategies and Random Mutation Hill-climbing (RMHC) In addition, algorithm pipeline which is specified by XML file is also provided for practitioner to configure & design evolutionary algorithms at ease. User can edit runtime & algorithm parameters in the configuration file (XML) & issue the command to execute "java -jar jaea.jar " Note that some standard EA configuration are provided in the executable package: GA (gaconf.xml), MA (maconf.xml) and PSO (psoconf.xml) [Less]

mpoems-in-java

Author: thomas.kremmel / gmail mPOEMS in JAVA is designed to provide optimisation problem developers with a tool to apply mPOEMS to their problem, without the need to know the complex algorithm in detail. Main goal is to use a state-of-the-art optimisation algorithm with small development effort. ... [More] Please find some more information about the project at the authors homepage http://thomaskremmel.com/mpoems/mpoems-info.php . This work was created at the Technical University of Vienna, in close collaboration with Jiri Kubalik from the Technical University of Prague and Stefan Biffl fromm the Technical University of Vienna. About mPOEMSCrossover, mutation, and evolution of a population are the basic principles of the evolution and natural selection. Evolutionary algorithms (EA) imitate the nature’s behaviour, and adapt the basic principles, to evolve a set of solutions, towards the optimum solution of a problem. mPOEMS is an optimisation framework, in the domain of evolutionary algorithms. It was designed to solve optimisation problems, with an unrestricted number of objectives. The algorithm has shown excellent performance in comparison with other state-of-the-art search algorithms. The paper including these results can be found using google scholar or directly using this link. The drawback of Evolutionary Algorithms is their complexity, and the development effort one has to conduct to create an implementation of such an algorithm. An impressing approach to test and compare various evolutionary algorithms, with little programming effort, is presented by the PISA project. The project website can be found by following this link. PISA is a platform and programming language independent interface for search algorithms. The use of PISA ensure easy, fast and reliable comparison of different optimization algorithms on various problems or benchmarks. With litte programming effort, one can create the problem specific part of an optimisation problem, and combine it with arbitrary ready-to-use, pre-compiled and very complex optimisation search algorithms. The project "mPOEMS in JAVA" shares the goal of providing an easy-to-use framework, to use complex search algorithms, with PISA. The PISA framework is applicable to almost any search algorithm. mPOEMS in JAVA focus on providing a framework to use the evolutionary algorithm mPOEMS, without the need to know the algorithm in detail. Getting StartedBest starting point is to read the documentation . The documentation is outdated (a bit) but nevertheless it`s the best starting point you can find ;-) Mainly it should be okay, but at least the settings part was not updated for a while. To see the framework in action just download the ready-to-start .jars at the downloads tab. There you can find two downloads. A mpoems-in-java solution for the knapsack problem and the .jar for the project selection problem. Get more information about the knapsack problem at wikipedia. With JAVA installed you can start the .jars with double-clicking on them and waiting for a while, checking out the resulting .xls after a few minutes or you can use this command to see the command-line output: java -jar tuwien.ifs.mpoems.knapsackProblem-1.0-SNAPSHOT-all.jar java -jar tuwien.ifs.mpoems.ppds.start-1.0-SNAPSHOT-all.jar Please note that you have to create an output folder for the ppds jar in order to create the output file. You can also change the settings mpoems-in-java is running with opening the .jar with win rar., change the settings file and then save it. You can find the settings file in the folder tuwien.ifs.mpoems.knapsackProblem-1.0-SNAPSHOT-all.jar\tuwien\ifs\mpoems\knapsackProblem\mPOEMSConf.cfg or tuwien.ifs.mpoems.ppds.start-1.0-SNAPSHOT-all.jar\tuwien\ifs\mpoems\ppds\start\mPOEMSConf.cfg tuwien.ifs.mpoems.ppds.start-1.0-SNAPSHOT-all.jar\tuwien\ifs\mpoems\ppds\start\projectSelection.cfg The download section also provides a .txt file with the most common maven commands. These commands should be enough to create an eclipse project out of the source code and get started. The JAVA implementation of mPOEMSmPOEMS in JAVA is the generic JAVA implementation of an evolutionary algorithm called Multiobjective Prototype Optimization with Evolved Improvement Steps (mPOEMS). The framework was designed to provide optimisation problem engineers with an interface to use mPOEMS, without detailed understanding of the algorithm. All needed methods and fields to use the framework are presented in the documentation of the framework which can be found at the following location: http://thomaskremmel.com/mpoems/mpoems_in_java_documentation.pdf . Main goal of the framework implementation was the creation of a generic framework, for which problem-dependent parts could be easily attached. The framework was created in close collaboration with Jiri Kubalik, the inventor of mPOEMS, under the observation of Stefan Biffl. About the authorsMag. Thomas Kremmel works as an expert for project management and information technology for aeon-group consulting. http://www.aeon-group.at and http://www.thomaskremmel.com Dr. Jiří Kubalík is assistant professor at the Department of Cybernetics at the Czech Technical University in Prague: http://cyber.felk.cvut.cz/people/page.php?id=38&detailed=y Prof. Dr. Stefan Biffl is associate professor of software engineering at the Institute of Software Technology and Interactive Systems, Vienna University of Technology. http://qse.ifs.tuwien.ac.at/~biffl/ About projects using the java implementation of mPOEMSThe thesis of the author served as a proof-of-concept for the framework. You can find the thesis here. Project Portfolio Decision Support Using Evolutionary Algorithms [Less]

ldse-algorithm

Low dimensional simplex evolution is a real-coded evolutionary algorithm for global optimization.

pyheuristics

General framework and many particular heuristics written in python. The collection includes genetic algorithms, evolutionary algorithm, artificial ant colony algorithm, genetic programming, swarm algorithms, artificial immune systems, harmony search, hyperheuristics and many others. Heuristics can be run on single machine or in parallel mode.

Evolutionary computation framework

ECF is a C++ framework intended for application of any type of evolutionary computation. Current features include: * parameterless: genotype (individual structure) is the only mandatory parameter * genetic algorithm genotypes (bitstring, binary encoded real values, floating point vectors ... [More] , permutation vectors), genetic programming genotype (tree) * individuals may contain any genotypes in any number * algorithms: steady state tournament, generational roulette-wheel, particle swarm optimization (PSO) * parallel execution in many models (global paralel EA, distributed EA, hybrid parallel EA...) using MPI [Less]