Computational Intelligence Library / News: Recent posts

Version 0.7.3 of CIlib has been released.

Additionally, the project downloads and hosting have been moved over to Google Code (http://code.google.com/p/cilib). Head on over to get the new release.

CIlib has moved it's source tree to Github.

Please go to: http://github.com/gpampara/cilib to see how to clone the repository and get the latest code. The SVN source tree will be disabled.

Sourceforge.net will still be the primary site for downloads etc.

The internal framework has a few limitations right now. As a result, CIlib will have some of the internal frameworks altered to be more generic and usable. One such framework is the Type framework.

A new release will be created as soon as this refactor has been completed.

With the approaching 0.7 release of CIlib, 0.6.11 represents a vital milestone in preparation for the 0.7 release.

Major documentation changes, bug fixes and optimisation have been added to the 0.6.11 release.

Specifics on the modifications are available within the changelog.

CILib 0.6.10 has been released with a large collection of fixes as well as some new functionality. This release can be regarded as a milestone for new functionality coming in the next few subsequent releases.

http://cilib.sourceforge.net

CILib 0.6.9 has been released. This release brings a correceted NichePSO implementation together with a large collection of refactorings and fixes.

http://cilib.sourceforge.net

CILib is a framework for developing Computational Intelligence software in swarm intelligence, evolutionary computing, neural networks, artificial immune systems, fuzzy logic and robotics.

This is a major feature update and bug fix release.

Please refer to the changelog for more information regarding the release.

CILib is a framework for developing Computational Intelligence software in swarm intelligence, evolutionary computing, neural networks, artificial immune systems, fuzzy logic and robotics.

This is a feature update and bug fix release.

the following additions have been made:

- Added initial code for the Multi-Objective Optimisation package (MOO)
- Added Sequance Alignment package in BioInformatics
- Added minor code and bug fixes
- Added generalised operators for Entities
- Added a new ControlParameterStrategy

CILib is a framework for developing Computational Intelligence software in swarm intelligence, evolutionary computing, neural networks, artificial immune systems, fuzzy logic and robotics.

This is a feature update and bug fix release. New classes have been added including a new fucntion and a bug fix on an existing function.

The addition of a the initial clustering framework has been made with a re-implementation of the NichePSO.... read more

CILib is a framework for developing Computational Intelligence software in swarm intelligence, evolutionary computing, neural networks, artificial immune systems, fuzzy logic and robotics.

This is a feature update release. The manner in which datasets are now handled is by far more descriptive and will scale better when more than one data set is used.

Minor bug fixes have also been made. These fixes, however, should remain transparent to all users.... read more

CILib is a framework for developing Computational Intelligence software in swarm intelligence, evolutionary computing, neural networks, artificial immune systems, fuzzy logic and robotics.

This version includes dataset support for all problems, if required, as well as the new skeleton framework for niching algorithms. The packages within the source tree have been refactored and renamed to conform to specifications. Many bug fixes and feature additions have been completed.... read more

CILib 0.5.2 has been released. This release includes the alteration of many of the packages within the source tree. These adaptions are all in preparation for a large collection of source that will be included (hopefully soon :-) ). Basic support has been added for EC topology structures and a division of the Algorithm classes has been commited :)

This version is a maintenance release that builds on the re-factored release of 0.5 (which was an unannounced and broken release). Notable improvements are a completely modularized PSO and the EC framework and PSO now finally share a common base interface.

This release may be better or worse than the previous release, but a release was long overdue, so here it is. Many potentially destabilising changes, including using Java 1.5 generics, refactoring and the testing of a new type system. Bug fixes to ACO and others. YMMV.

This release features minor changes to the domain package API (vectors are now flattened), the addition of the domain visitor interface, minor bug fixes, added unit tests, ant colony updates, a start to refactoring the cooperative optimisation algorithm and updated documentation.

This release fixes a nasty VonNeumannTopology divide by zero bug.

This is a minor bugfix release.

Highlights of this release include: more updated documentation; a new domain architecture allowing for more than just continous domained problems; ant colony optimisation; early games framework; tree and graph container classes; more benchmark functions; general architectural cleanup; and bug fixes.

This is a long overdue release, they should become more frequent from now on. Highlights include: a new PSO architecture, resizeable PSO topologies, NichePSO improvements, a new LFPSO algorithm, new benchmark functions, bug fixes and some initial evolutionary algorithms and neural network work.

The Computational Intelligence Research Group at the University of Pretoria (http://cirg.cs.up.ac.za) has opted to standardise on CILib. All future development done by new members joining CIRG@UP
will be done within the CILib framework. Hopefully, this will jump-start development of CILib in CI paradigms other than particle swarms.

CILIB 0.2 includes an implementation of NichePSO. The NichePSO code has been contributed by Clive Naicker. Clive is now a member of the CILib development team.

CILib (Computational Intelligence Library) is a framework for
computational intelligence algorithms. This is the initial release which includes a basic simulation engine (for measuring algorithm performance), a number of particle swarm optimisers and some limited documentation.