超级会员免费看
本文提出了一种无监督强化学习代理，该代理通过学习离散的像素分组模型来实现结构化的探索。这种方法允许代理在保持环境几何形状的同时，通过内在奖励函数（如质心坐标和面积）学习控制策略。这些策略形成行为基础，用于一致的探索，并在分层强化学习框架中解决外在奖励问题。实验表明，该方法在3D导航和具有稀疏奖励的Atari游戏中表现出竞争力。
LEARNING TO CONTROL VISUAL ABSTRACTIONS FOR STRUCTURED EXPLORATION IN DEEP REINFORCEMENT LEARNING
ABSTRACT
Exploration in environments with sparse rewards is a key challenge for reinforcement learning. How do we design agents with generic inductive biases so that they can explore in a consistent manner instead of just using local exploration schemes like epsilon-greedy? We propose an unsupervised reinforcement learning agent which learns a discrete pixel grouping model that preserves spatial geometry of the sensors and implicitly of the environment as well. We use this representation to derive geometric intrinsic reward functions,
订阅专栏 解锁全文
扫一扫
1337
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
