Alyuda NeuroFusion is a general-purpose neural networks library that can be used to create, train and apply constructive neural networks for solving both regression and classification problems.
Is a neural networks library written in ANSI C++ and compiled with Visual C++ 6.0. With this library you can create, train and apply constructive neural networks for both regression and classification problems.
Enhance your software
Artificial Intelligence provides key functionality that improves customer satisfaction, puts your software ahead of competition and distinguishes it from similar products.
Save development time
All theoretical information is hidden inside the library. You do not have to tweak with training parameters and experiment with different architectures, activation functions, stopping conditions, etc. Your development time is reduced significantly due to the fact that you have to deal with a minimum set of functions.
NeuroFusion is not an old-fashioned back-propagation. These are state-of-the-art constructive neural networks combined with proprietary Alyuda algorithms of automatic data preprocessing, network design and training.
The library does not require knowledge in neural networks and statistics. Simply feed your data, call train function and read the results. NeuroFusion does all math calculations automatically. It preprocesses and partitions your data, initializes neural networks and even intelligently selects stopping conditions and controls generalization loss to provide you with the best solution neural networks can produce.
If you have knowledge in neural networks, you can control the training process using real time training statistics, modify stopping conditions, and specify data partition parameters. |