资 源 简 介
Java implementation of a Feed Forward Neural Network (FFNN).
The design is fairly object orientated which means there is a trade off between performance and code readability.
The design has had a focus on experimentation, therefore it is easily configured to different network sizes, number of layers. There are also two activation functions: Logistic (default) and the Hyperbolic Tangent. There are two training methods Online Back Propagation, and Particle Swarm Optimization.