SNIPE (click here to get on the dedicated page) is a well-documented JAVA library that implements a framework for neural networks in a speedy, feature-rich and usable way. It was originally designed for high performance simulations with lots and lots of neural networks (even large ones) being trained simultaneously, and therefore it is pretty optimized. Have fun with it!

SNIPE is designed with respect to each of the following goals:

• Generalized data structure for arbitrary network topologies, so that virtually all network structures can be realized or even easily hand-crafted.
• Built-In, fast and easy-to-use learning operators for gradient descent or evolutionary learning, as well as mechanisms for efficient control of even large network populations.
• Mechanisms for design and control of even large populations of neural networks
• Optimal speed neural network data propagation in contrast to naive data structures, even in special cases like multilayer perceptrons and sparse networks, as well as low computational topology editing effort.
• Low memory consumption - grows only with the number of synapses, not quadratically with the number of neurons
• In-situ processing - no extra memory or preprocessing of the data structure is necessary in order to use the network after editing
• Usage of only low level data structures (arrays) for easy portability. It is not the goal to quench the last tiniest bit of asymptotic complexity out of the structure, but to make it usable, light weight and fast in praxis.
• No object-oriented overload, like objects for every neuron or even synapses, etc.

This web site is migrated on a new server, so there may be some small offline periods around Sunday, March 7th, sorry for that.

New Layout, lots of enhancements, new content and of course bilingual in German and English.

As usual, the manuscript can be downloaded at the subpage about Neural Networks. In the epsilon version, the technical base of the manuscript was completely renewed. In the text, there are lots of updates, too.

• New Chapter: Biological Neural Networks
• New layout and design, which makes use of common typography rules. Relayouted type area and spaces.
• New page headers and footers, containing section and chapter titles.
• New, modern fonts. Not that you'll see lots of difference, but they can be read easier.
• “Speaking headlines” througout the entire manuscript boil down the material of the following section to one sharp sentence.
• Clickable links and pdf-contained table of contents throughout the entire manuscript.
• Whole lots of little enhancements and corrections that I was advised to by readers. Thanks a lot, guys! In this context, the list of helpful readers in the preface was extended thoroughly
• Removed the two Table of contents and replaced them by one single TOC of intermediate lenght.
• Several rearrangements, in particular parts of the perceptron chapter (5) were moved towards the learning chapter (4). The Regional and Online Learnable Fields Chapter was moved into the Excursus about Clustering. The Part “Further Views on Learning” was removed, its chapters were rearranged.
• New CC-license, see above
• Cool new cover

A very, very big “thank you” goes to Beate Kuhl again for the translation!!!