See the RobocodeJGAP page for details.
JGAP has been subject of various publications, applications and research projects. Find a list of currently known uses of JGAP below in loose order. If you know any other to be included here, please tell us (see main page, feedback section for contact information)!
If you want to cite JGAP, please do it as follows: Meffert, Klaus et al.: JGAP - Java Genetic Algorithms and Genetic Programming Package. URL: http://jgap.sf.net
This dissertation describes a newly developed Adaptive Traffic Control System for urban road networks with several signalized intersections. Traffic signal control at all intersections is continuously adapted to the currently estimated traffic demand. In this context JGAP has been used to optimize the coordination of these intersections in order to reduce waiting and travel times.
Author: Tobias Pohlmann
Subject is the evolution of a game playing strategy for Backgammon. To accomplish this, the neural network framework JOONE has been used in conjunction with JGAP. The bridge between the two packages is another Sourceforge project named JOONEGAP.
Author: Michael James Collins
The research involves extending JGAP in order to
model genetic flow within metapopulations. The results are used to
investigate the relationship between population connectivity, population
dynamics and population genetic structure.
Author: Jonathan Kool
This research involves the use of JGAP in developing an Evolutionary Algorithm to assign airport arrival and departure times to flights based on a multi-objective fitness function. The fitness function includes historic flight and taxi times as well as airline and air traffic control preferences. To the abstract.
Author: Florian Hafner
JGAP has been used to help develop database performant queries for a huge number of datasets.
Author: Achim Westermann
Original title in German: Integration Evolutionärer Algorithmen in das MFOS
The thesis considers the following aspects:
Exploration of optimizing Fuzzy Controllers with Evolutionary Algorithms
Development of a Genetic Algorithm (GA) for optimizing the rule-base of Mamadani-type Fuzzy Controllers following the Pittsburgh-Approach
Design and Implementation of the Integration of the developed GA in the existing Java-Implementation of MFOS for Mamadani-type Fuzzy Controllers (MFOS-M)
Experiments with the Implementation of the GA show good performance
Author: Knut Willems
Title of the paper: Impact of Energy End-use and Customer Interruption Cost on Optimal Allocation of Switchgear in Constrained Distribution Networks
Accepted for future publication in IEEE Transactions on Power Delivery, see http://ieeexplore.ieee.org/.
Abstract: The introduction of new energy carriers, such as natural gas and district heating, to energy systems dominated by electrical power will certainly relieve stress on the power system. Some of the end uses initially served by the power system will be gradually decoupled and served by alternative energy carriers. As a result, the specific customer interruption costs and load profiles will change. In this paper, we analyze how the optimal level of switchgear in electric power distribution systems is affected by such changes. The proposed optimization method is based on a genetic algorithm and takes into account the constrained network capacity.
Authors: A. Helseth; A. T. Holen
Abstract: Collaborative work appears between intelligent agents of different types. to make independent groups of workers from some categories, like carpenters, brick layers, etc. To discover their collaborative attitudes they use the scoring method, where every worker scores the others from different trades.The objectives are to form groups of workers with high compatibility value and to have a high compatibility value for the worst group, too. The problem becomes more interesting if software collaborative groups or specialized intelligent agents are involved. One has to prospect also the level of knowledge overlap between the trade groups of agents. This paper resumes to the problem of construction workers so as there is no overlap between the trades and the level of knowledge is not in the universe of discourse. We propose a Greedy and a genetic algorithm approach and we compare these methods.
Author: László Illyés
Abstract: This paper and presentation was submitted by
Mark McFadden for a Data Communication and Networking graduate level class at
About the author: Mark McFadden is a software developer
that works for a large financial institution in the
Download the paper and the presentation
Abstract: This paper investigates the use of genetic algorithms in test data generation from state-based systems. For chosen paths in the state machine diagram (sequence of method calls) the input parameters are automatically generated using genetic algorithms and JGAP, in order to satisfy the transition constraints or guards.
Appeared in: Proceedings of the Ninth International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC 2007), IEEE Computer Society, pp. 188-195, 2007
Authors: Raluca Lefticaru, Florentin Ipate
Abstract: In this paper an approximate graph-based model of the system under test is built using a genetic algorithm implemented with JGAP. The obtained model is further used as basis for test generation.
Appeared in: Romanian Journal of Information Science and Technology, 11(3), pp. 209–227, 2008
Authors: Florentin Ipate, Raluca Lefticaru
Abstract: The broadcasting algorithm is specified in this paper using P systems, with different rules formats. The unknown elements of the P system are determined using genetic algorithms implemented with the JGAP package.
Appeared in: Proceedings of the Tenth Workshop on Membrane Computing (WMC10), pp. 337-354, 2009
Authors: Raluca Lefticaru, Florentin Ipate, Marian Gheorghe, Gexiang Zhang
Abstract: This paper addresses the problem of automatic model generation, using model checking and genetic algorithms. It proposes and empirically evaluates a new type of fitness function which, by taking also into account the counterexamples provided by a model checker, improves the success rate of the genetic algorithm. The implementation was done with JGAP.
To appear in: Proceedings of the 4th Balkan Conference in Informatics (BCI'09), IEEE Computer Society
Authors: Raluca Lefticaru, Florentin Ipate, Cristina Tudose
Abstract: The objective of this study was to improve a herbage growth model. The existing and new versions of the model were each calibrated, using JGAP (minimizing the sum of square error between predicted and observed), against five time series of herbage growth data using the first half of each time series. The second half of each time series was used for validation. A suite of goodness-of-fit indicators was used to evaluate and compare the existing and new versions of the model.
Appeared in: New Zealand Journal of Agricultural Research 52, pp 477-494
Authors: A.J. Romera, D.G. McCall, J.M. Lee, M.G. Agnusdei
Download the paper or see here
Abstract: Knowing the amount of herbage mass
available on the farm (ideally measured weekly) is an important step in
achieving high pasture utilization on pastoral dairy farms in New Zealand, but
the information must be used in a timely manner to make efficient management
decisions. However, most New Zealand dairy farmers do not measure their pastures
regularly. This project aimed to develop a simple alternative, in the form of a
prototype software tool (Pasture Growth Simulation Using Smalltalk, PGSUS) to
predict herbage mass at an individual paddock level, which reduces (not
eliminates) the requirement for physical data collection and provides more
information from the measurements taken. The main data requirements are pasture
herbage mass for each paddock and grazing or cutting events. A climate-driven
pasture simulation model is used to predict herbage mass between intermittent
pasture measurements. The pasture model contains certain empirical parameters
that are fitted to the observed data for each paddock individually, using all
the previous data to “train” the model. PGSUS requires daily weather data,
including mean, minimum and maximum air temperature, solar radiation, rain and
potential evapotranspiration. Data from the Virtual Climate Station Network
(VCSN) from NIWA (National Institute of Water and Atmospheric Research Ltd., New
Zealand) are used to drive the model. Preliminary testing was done on two
commercial dairy farms, one in the Waikato (North Island) and the other in the
Canterbury (South Island) regions of New Zealand. Up to 28 days without
measurements, PGSUS estimated herbage mass with correlation of approximately 0.9
and with small bias.
Appeared in: Computers and electronics in agriculture, Volume 74, issue 1, October 2010, pp. 66-72
Authors: A.J.
Romera, P. Beukes, C. Clark, D. Clark, H. Levy and A. Tait
Comparison of several evolutionary parameters for JGAP, done in the context of a thesis (last update: 17 August 2005). Besides an in-depth analysis, the article contains some nice graphs.
Author: Rebecca Fiebrink
German book "JUnit Profi-Tipps", describing software testing with JUnit, and referencing JGAP as a prominent project massively utilizing unit tests. Link to Amazon.
Author: Klaus Meffert
Printed article in the German magazine JavaMagazin 08/2004. Original title: "Auf Darwins Spuren". Link to number 08/2004.
Author: Klaus Meffert
Printed article in the German magazine Der Entwickler 02/2005. Original title: "Einführung in die Genetische Programmierung". Link to number 02/2005. Download the article (local version)
Author: Klaus Meffert
Printed article in the German magazine JavaMagazin 05/2005. Original title: "Genetische Programmierung: leistungsfähiges Verfahren zur Problemlösung". Link to number 05/2005.
Author: Klaus Meffert
Weblog entry in the popular SAP Developer Network (SDN): Genetic Algorithms: Introducing JGAP
Author: Klaus Meffert
Weblog entry in the popular SAP Developer Network (SDN): Genetic Algorithms: The Darwinian Way
Author: Klaus Meffert
In Self-Managing Federated Services, Cuenca-Acuna et. al. use JGAP to find mappings of processes to nodes on a distributed system that optimize a Service-Level Agreement (SLA).
Authors: Francisco Matias Cuenca-Acuna, Thu D. Nguyen
Please see http://research.microsoft.com/~livshits/papers/pdf/thesis.pdf.
Author: Benjamin Livshits
A work on automatic human-readable melody characterization. Utilizes JGAP to implement GA experiments. Please see http://grfia.dlsi.ua.es/repositori/grfia/pubs/212/smc-08.pdf
Authors: Pedro J. Ponce de León, David Rizo, Rafael Ramirez, José M. Iñesta
Seems related to the previous reference. Gives a description of the steps necessary to do a melody composition with JGAP, along with source code snippets: Jamie Craane's Blog.
Author: Jamie Craane, June 16, 2009
|
|
See the RobocodeJGAP page for details. |
A Starcraft 2 Genetic Algorithm (Build Order Search).
JOONEGAP allows for adjusting a Neural Network with help of a Genetic Algorithm. The frameworks involved are JOONE and JGAP.
Attention: developers needed for bringing the project up-to-date! Please contact us in case you are interested.
Result of the above mentioned dissertation is a full-blown Backgammon game with AI opponent, game board and nice-to-use graphical gaming interface. You could find it within JOONEGAP.
unEvo is an Eclipse plug-in designed to be a framework for the implementation of Evolutionary Algorithms focused on the Experimentation and Research process rather than the codification of the algorithm itself. EvoGAP in turn is a bridge between unEvo and JGAP. EvoGAP is currently under initial development.
HiperionCAD is a CAD tool for optical fiber networks design with a genetic algorithm optimization approach. Its a project sponsored by the North Electrical Company of Brazil's Eletronorte, and it was built using Eclipse, its Graphical Modeling Framework and JGAP. In a rough way, JGAP its used to model all the parameters of network components onto genes, so it can be found out find very quickly which combination of values are the less expensive (in terms of wasted signal strength) on the network.
This is a really nice project as you can imagine from the screenshots:
To obtain more information please visit the HiperionCAD website or contact
Ugo Sangiorgi - Chief on development
Diego
Guimarães - Developer
Paulo Augusto Bichara - Developer
and GA Specialist
See this blog entry from Berlin Brown for a JGAP port to the programming language Scala. Source code included. Or download a PDF file including the dump of the web page.
DOAP contains all general meta information about JGAP, including links to resources.
No need to download something in case you want to get a quick overview from anywhere in the online-world. Just browse the trunk of JGAP files at Koders.
Copyright © 2005-2012 Klaus Meffert, GNU free documentation license