Efficient Contextual Bandits with Continuous Actions
This work addresses the challenge of efficient decision-making in continuous action spaces for applications like online advertising or recommendation systems, representing an incremental improvement by building on existing supervised learning frameworks.
The paper tackles the problem of designing a computationally tractable algorithm for contextual bandits with continuous actions, achieving this through a reduction-style method that composes with most supervised learning representations and is validated with large-scale experiments.
We create a computationally tractable algorithm for contextual bandits with continuous actions having unknown structure. Our reduction-style algorithm composes with most supervised learning representations. We prove that it works in a general sense and verify the new functionality with large-scale experiments.