This plugin implements the Machine Learning library Tensorflow.NET in Grasshopper to provide some Machine Learning functionalities in Grasshopper. Additional components to analyze problems related to circularity of building materials are included. The included example files show how to define Supervised Learning problems, how to save and load trained networks, how to use the Keras API in Grasshopper, GANs and how to tackle Deep Reinforcement Learning in Grasshopper.
The installation of the plugin "Aneome" is also required to define Reinforcement Learning problems in Grasshopper:
"Owl" can be also used for tensor visualization:
Special thanks to Björn Pettersson for the support with the development of the DQN agent.