TorontoMetropolitanUniversity/Network_Defense_Symmetric_Competitive
收藏资源简介:
该数据集包含102,400,000个时间步,专注于多智能体强化学习场景。红方智能体的目标是发现漏洞、提升权限、破坏资产并维持持久性,其行动空间基于MITRE ATT&CK框架的阶段建模。蓝方智能体的目标是维护系统可用性、减少攻击面、检测恶意行为、通过蜜罐欺骗攻击者并驱逐入侵者,其行动空间包括防御配置和缓解措施。双方呈现纯零和对立关系,即红方的收益是蓝方的绝对损失,反之亦然。
102,400,000 timesteps, Multi-Agent Reinforcement Learning. The Red Agent’s goal is to discover vulnerabilities, elevate privileges, compromise assets, and maintain persistence, with its action space modeled after phases of the MITRE ATT&CK framework. The Blue Agent’s goal is to maintain system availability, reduce the attack surface, detect malicious behavior, deceive the attacker with honeypots, and evict the intruder, with its action space consisting of defensive configurations and mitigation steps. Pure Zero-Sum Opposition: Reds gain is Blues absolute loss and vice versa.




