the function generate_wasserstein_reward and the variable fisher

cvigoe / DRL4MAAS

Code for paper "Multi-Agent Active Search: a Reinforcement Learning Approach", submitted to ICRA 2022.

MIT License

12 stars 0 forks source link

the function generate_wasserstein_reward and the variable fisher #2

Open chengyh23 opened 2 years ago

chengyh23 commented 2 years ago

what does the variable fisher in function generate_belief_rep (e.g., in agnosticmaas_env.py) mean?

chengyh23 commented 2 years ago

According to the paper, "to provide more learning signal to the agent", wasserstein distance is used as reward. However, in the real-world scene, is there any way of interacting with the environment, so that the agent can get a feedback which reflect the Wasserstein distence between distribution of the belief and the state beneath it?