Knowledge Transfer for Deep Reinforcement Learning with Hierarchical Experience Replay
نویسندگان
چکیده
The process for transferring knowledge of multiple reinforcement learning policies into a single multi-task policy via distillation technique is known as policy distillation. When policy distillation is under a deep reinforcement learning setting, due to the giant parameter size and the huge state space for each task domain, it requires extensive computational efforts to train the multi-task policy network. In this paper, we propose a new policy distillation architecture for deep reinforcement learning, where we assume that each task uses its taskspecific high-level convolutional features as the inputs to the multi-task policy network. Furthermore, we propose a new sampling framework termed hierarchical prioritized experience replay to selectively choose experiences from the replay memories of each task domain to perform learning on the network. With the above two attempts, we aim to accelerate the learning of the multi-task policy network while guaranteeing a good performance. We use Atari 2600 games as testing environment to demonstrate the efficiency and effectiveness of our proposed solution for policy distillation.
منابع مشابه
Hierarchical Functional Concepts for Knowledge Transfer among Reinforcement Learning Agents
This article introduces the notions of functional space and concept as a way of knowledge representation and abstraction for Reinforcement Learning agents. These definitions are used as a tool of knowledge transfer among agents. The agents are assumed to be heterogeneous; they have different state spaces but share a same dynamic, reward and action space. In other words, the agents are assumed t...
متن کاملPrioritized Experience Replay
Experience replay lets online reinforcement learning agents remember and reuse experiences from the past. In prior work, experience transitions were uniformly sampled from a replay memory. However, this approach simply replays transitions at the same frequency that they were originally experienced, regardless of their significance. In this paper we develop a framework for prioritizing experienc...
متن کاملA Deeper Look at Experience Replay
Experience replay plays an important role in the success of deep reinforcement learning (RL) by helping stabilize the neural networks. It has become a new norm in deep RL algorithms. In this paper, however, we showcase that varying the size of the experience replay buffer can hurt the performance even in very simple tasks. The size of the replay buffer is actually a hyper-parameter which needs ...
متن کاملStabilising Experience Replay for Deep Multi-Agent Reinforcement Learning
Many real-world problems, such as network packet routing and urban traffic control, are naturally modeled as multi-agent reinforcement learning (RL) problems. However, existing multi-agent RL methods typically scale poorly in the problem size. Therefore, a key challenge is to translate the success of deep learning on singleagent RL to the multi-agent setting. A key stumbling block is that indep...
متن کامل