Neuroph Manifesto

April, 2014.

Neuroph Project has gone a long way, from its first public release at 2008. to Dukes Choice Award at 2013. It has become a succesfull open source project thanks to its users and contributors, and one of the world's leading Java neural network frameworks. It is being used by developers, students and researchers all around the world.

The purpose of this manifesto is to announce the goals and plans for the future Neuroph development in order to encourage interested parties to get involved, and contribute.

Future Development Plan

Future development will include the following:

1. New Neuroph framework features

• Support for all known artificial neural network architectures and learning rules based on common framework

• Deep learning algorithms

• Performance improvements and support for parallel processing (multicore, GPU, HPC)

• Spiking networks

• Genetic algorithms, fractal and modular networks

• Integrated development, simulation, analysis and visualization tools

2. Neural Network Applications

• Image recognition

• Medical Image Recognition

• Face recognition

• Optical Character Recognition

• BCI (EEG signal processing)

• Brain Research and Simulations

• (open for all other applications received from community)

3. Community Building

• Create official community programmes, so institutions and individuals can participate in development and other activities.

• Promote projects that use Neuroph and provide higher level of support for projects registered in community programmes.

• Support and promote educational institutions that use Neuroph and provide higher level of support for those registered in community programmes.

• Include community members to provide support, track bugs, feature requests, and fixes.

• Support contributions from community (provide documentation and mentoring for interested contributors)

• Collaboration and integration with other related open source projects.

Mission

The misison of the Neuroph project it to provide the state of the art, open source, platform independent software framework and tools for neural network development, research and education.

Core Team and Contributors

Neuroph core development team is located at University of Belgrade, Faculty of Organisational Sciences, Serbia, within Laboratory for Artificial intelligence and Open Source Software Development Center. Neuroph is being used for teaching neural networks and software development at University of Belgrade, and each year we have several students that contribute through their working assignements, graduate or master thesis. Beside that we get contributions from universities and individuals from all over the world.

Collaboration

Neuroph team is opened for collaboration with individuals, universities, companies or other organisations that have interest in Neuroph project. Neuroph team can participate as a partner in research or commercial projects, to provide software development, research or support services.

For more info see our TODO list and how you can contribute

Dr Zoran Sevarac
Founder of the Neuroph project
University of Belgrade, Faculty of Organisational Sciences
Member of the
NetBeans Dream Team
Java Champions
GOOD OLD AI Network