… In lifelong reinforcement learning, agents must effectively transfer … learning. To this end, we
here develop theory to compute and use state abstractions in lifelong reinforcement learning. …

… The lifelong reinforcement learning (RL) setting formalizes the problem of building agents that
must … The key question in lifelong RL is the question of transfer: how can algorithms exploit …

… algorithm for lifelong policy search in reinforcement learning. … reinforcement learning instead
of lifelong learning setting. … lifelong policy gradient RL algorithm that can efficiently learn …

… an agent is facing a series of Reinforcement Learning (RL) tasks. … to a value-transfer method
for Lifelong RL, which we use to build … We illustrate the benefits of the method in Lifelong RL …

… the novel problem of lifelong learning from demonstration, … demonstrated tasks to accelerate
the learning of new tasks, … lifelong learning approach to inverse reinforcement learning, …

… of how a reinforcement learning agent that … lifelong learning: when faced with a sequence
of MDPs sampled from a distribution over MDPs, how can a reinforcement learning agent learn …

… reinforcement learning framework to real-world problems. For that purpose, we think that
the \lifelong… This paper presented a reinforcement learning approach dealing with multiple-tasks…

… The core contribution of this work is a framework for efficient lifelong reinforcement learning
… Our simulated experiments consider two lifelong RL problems: a sequence of key-insertion …

… during lifelong learning across a series of RL tasks that is guaranteed to reduce or match …
-task learning in future RL tasks. Third, we present a two-phase lifelong learning algorithm that …

… progress in reinforcement learning for robotics, gameplay, … reinforcement learning to the
evolving, open-world problems often found in critical application spaces. Reinforcement learning …