arXiv (Cornell University)· 2016· Preprint
Continuous control with deep reinforcement learning
- 6,787citations
- 2016year
Short summary
A novel actor-critic, model-free algorithm, the Deterministic Policy Gradient (DPG), robustly solves over 20 simulated physics tasks with continuous action spaces, matching planning algorithm performance and learning end-to-end from raw pixels.
AI-generated from the title and abstract; the full text is not read.
Key points
- Introduced a model-free, actor-critic algorithm using deterministic policy gradients for continuous action spaces.
- Achieved robust performance across 20+ simulated physics tasks (e.g., cartpole, manipulation, locomotion, driving).
- Algorithm performance matched planning algorithms with full domain dynamics access.
- Demonstrated end-to-end learning directly from raw pixel inputs for many tasks.
AI-generated from the title and abstract; the full text is not read.
Abstract
Abstract: We adapt the ideas underlying the success of Deep Q-Learning to the continuous action domain. We present an actor-critic, model-free algorithm based on the deterministic policy gradient that can operate over continuous action spaces. Using the same learning algorithm, network architecture and hyper-parameters, our algorithm robustly solves more than 20 simulated physics tasks, including classic problems such as cartpole swing-up, dexterous manipulation, legged locomotion and car driving. Our algorithm is able to find policies whose performance is competitive with those found by a planning algorithm with full access to the dynamics of the domain and its derivatives. We further demonstrate that for many of the tasks the algorithm can learn policies end-to-end: directly from raw pixel inputs.
The authors' abstract, as published at the source. arXiv (Cornell University), 2016 · DOI ↗
Continue with a free account
Ask the paper: 3 free questions a day about this paper; save it, get its citation, new summaries every day for your field. Takeaways are Premium.
Continue free on the webSign in with Google or Apple; no card needed. You come back to this paper.
On your phone:
Field: Artificial Intelligence
Artificial IntelligenceComputer Science