A library provides integration between Domain Knowledge and Deep Learning.
Project description
The library allows to specify a problem domain with a conceptual [graph](https://github.com/HLR/DomiKnowS/blob/main/docs/developer/KNOWLEDGE.md#graph) including declarations of edges and nodes, as well as [logical constraints](https://github.com/HLR/DomiKnowS/blob/main/docs/developer/KNOWLEDGE.md#constraints) on the graph concepts and relations. [Neural network] (https://github.com/HLR/DomiKnowS/blob/main/docs/developer/MODEL.md#model-declaration) outputs bounded to the graph edges and nodes. The logical constraints are conveted to ILP and Gurobi Solver is used for inferencing. This adds a relational overlay over elements in a network that relates physical concepts in applications. <br> The example runing in Google colab environment, presenting the usage of the libarary is [here](https://colab.research.google.com/drive/1FvdePHv3h3NDSTkBw1VKwAmaZFWuGgTi)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.