Gym-μRTS Environment

该数据集采用了Gym-μRTS环境，用于在具有变化动作的环境中训练和评估智能体，并与预设的对手进行对抗。该数据集包含了智能体在与13个预设AI对手进行评估时的性能指标，具体的细分数据在论文中有详细阐述。此外，该数据集还涉及了在8x8和16x16地图上训练的两个模型，其任务是强化学习。

This dataset utilizes the Gym-μRTS environment, which is designed for training and evaluating AI Agents in environments with variable action spaces, as well as conducting adversarial matches against pre-defined opponents. The dataset encompasses performance metrics of AI Agents when evaluated against 13 pre-set AI opponents, with detailed breakdowns of the data presented in the associated paper. Furthermore, this dataset includes two models trained on 8x8 and 16x16 maps, which are developed for reinforcement learning tasks.

搜集汇总

背景与挑战

背景概述

该数据集基于Gym-μRTS环境，专为强化学习设计，用于训练和评估智能体在动态动作环境中对抗13个预设AI对手的性能。数据集覆盖8x8和16x16两种地图规模，具体性能指标和细分数据在关联论文中提供，适用于研究智能体在复杂对抗场景中的适应能力。