Simulates MLP, RMLP and Cascade-Correlation models with dynamic size adjustment algorithms. Includes various training patterns, error and activation functions.
Java open source neural network framework which can be used to create and train common types of neural networks. Site contains full source code, documentation and project info.
An open source library for numeric computation using data flow graphs, optimized for parallel processing using GPUs to handle massive scale deep neural network training. Offers documentation, tutorials and source code.
Neural network library implemented in ANSI C with bindings for 20 other languages. Creates multilayer feedforward networks with support for both fully connected and sparse connected networks.
A Python neural network system built on top of Theano which provides transparent multi-core and GPU optimization for efficiently training multilayer convolutional neural networks.