sarlopt uses single-agent Reinforcement Learning (RL) and Deep Reinforcement Learning (Deep RL) to learn optimization algorithms automatically. We intend to convert the repo to a library in the future. Currently, the codes are implemented in Python with TensorFlow and TF-Agents.
- marlopt: Multi-Agent Reinforcement Learning for Global Optimization.
- optfunc: Test functions for global optimization algorithms.
- TensorFlow
- TF-Agents