论文标题
通过计划准米数的多任务增强学习
Multi-task Reinforcement Learning with a Planning Quasi-Metric
论文作者
论文摘要
我们介绍了一种新的强化学习方法，结合了计划准米特（PQM），该方法估计了从任何状态到另一个状态所需的步骤数，而特定于任务的“瞄准器”计算了目标状态以达到给定目标。这种分解允许跨任务的跨任务无关模型模型共享，该模型可以捕获环境的动态，并且可以以密集且无监督的方式学习。与最近发表的有关标准位叉问题以及在Mujoco机器人臂模拟器中的方法相比，我们实现了多重训练的速度。
We introduce a new reinforcement learning approach combining a planning quasi-metric (PQM) that estimates the number of steps required to go from any state to another, with task-specific "aimers" that compute a target state to reach a given goal. This decomposition allows the sharing across tasks of a task-agnostic model of the quasi-metric that captures the environment's dynamics and can be learned in a dense and unsupervised manner. We achieve multiple-fold training speed-up compared to recently published methods on the standard bit-flip problem and in the MuJoCo robotic arm simulator.
