How to train one agent and don't update another one in multi-agent reinforcement learning setup?

1 回表示 (過去 30 日間)
Lingfeng Tao
Lingfeng Tao 2021 年 7 月 30 日
Hi, I have a multi-agent reinforcment learning environment with two agents. One agent is pretrained (agentA) and the perofrmance is good. I need to train another one (agentB) to cooperate with agentA. I'm wondering if I can only train agentB and keep agentA unchanged so its performance will not be influenced.
Thanks!

サインインしてこの質問に回答する。

回答 (0 件)

カテゴリ

Help Center および File Exchange で Reinforcement Learning についてさらに検索

タグ

製品


リリース

R2021a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by