C# libraries written written to model different neural networks, and deep learning algorithms that train them.
/badge.svg){kind=icon}
The following neural networks are supported:
| Network | Supports |
|---|---|
| Dense Neural Networks | Deep networks |
| Convolutional Neural Networks | 1D + filtering, 2D + filtering |
The following activation functions are supported:
There is an implementation of backpropagation and negative sampling (both with momentum), where the following error functions are supported:
| Function | Links |
|---|---|
| Mean Squared Error | https://en.wikipedia.org/wiki/Mean_squared_error |
| Cross Entropy | https://en.wikipedia.org/wiki/Cross_entropy |
There is an implementation of Word2vec, supporting both Skip Gram and CBOW architectures. To analyse word embeddings, there are implementations of the following algorithms:
| Algorithm | Links |
|---|---|
| k-means clustering | https://en.wikipedia.org/wiki/K-means_clustering |
| Density-based spatial clustering of applications with noise (DBSCAN) | https://en.wikipedia.org/wiki/DBSCAN |
| T-distributed Stochastic Neighbor Embedding (t-SNE) | https://en.wikipedia.org/wiki/T-distributed_stochastic_neighbor_embedding |
Open the GingerbreadAI.NeuralNetwork.Test solution file.
Visual Studio 2019 (with dotnet core 3.1).
All unit tests should be runnable from the GingerbreadAI solution.
Longer running tests that produce a report need to be run in debug mode, and can be found in the GinbergreadAI.NeuralNetwork.Test project (in the GingerbreadAI solution).
See the list of contributors who participated in this project.
See our Contributing Guidelines.
See our Code of Conduct.
This project is licensed under the MIT License - see the LICENSE file for details.