Simplifying Reward Design through Divide-and-Conquer
June 07, 2018 Β· Declared Dead Β· π Robotics: Science and Systems
"No code URL or promise found in abstract"
Evidence collected by the PWNC Scanner
Authors
Ellis Ratner, Dylan Hadfield-Menell, Anca D. Dragan
arXiv ID
1806.02501
Category
cs.RO: Robotics
Cross-listed
cs.AI,
cs.LG
Citations
31
Venue
Robotics: Science and Systems
Last Checked
3 months ago
Abstract
Designing a good reward function is essential to robot planning and reinforcement learning, but it can also be challenging and frustrating. The reward needs to work across multiple different environments, and that often requires many iterations of tuning. We introduce a novel divide-and-conquer approach that enables the designer to specify a reward separately for each environment. By treating these separate reward functions as observations about the underlying true reward, we derive an approach to infer a common reward across all environments. We conduct user studies in an abstract grid world domain and in a motion planning domain for a 7-DOF manipulator that measure user effort and solution quality. We show that our method is faster, easier to use, and produces a higher quality solution than the typical method of designing a reward jointly across all environments. We additionally conduct a series of experiments that measure the sensitivity of these results to different properties of the reward design task, such as the number of environments, the number of feasible solutions per environment, and the fraction of the total features that vary within each environment. We find that independent reward design outperforms the standard, joint, reward design process but works best when the design problem can be divided into simpler subproblems.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
π Similar Papers
In the same crypt β Robotics
R.I.P.
π»
Ghosted
R.I.P.
π»
Ghosted
AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles
π
π
The Cartographer
A Survey of Motion Planning and Control Techniques for Self-driving Urban Vehicles
π
π
The Cartographer
Unmanned Aerial Vehicles: A Survey on Civil Applications and Key Research Challenges
π
π
The Cartographer
A Survey of Autonomous Driving: Common Practices and Emerging Technologies
R.I.P.
π»
Ghosted
Learning agile and dynamic motor skills for legged robots
Died the same way β π» Ghosted
R.I.P.
π»
Ghosted
Federated Learning: Strategies for Improving Communication Efficiency
R.I.P.
π»
Ghosted
In-Datacenter Performance Analysis of a Tensor Processing Unit
R.I.P.
π»
Ghosted
Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning
R.I.P.
π»
Ghosted