Files and links (1)
Preprint (Author's original)arXiv.org - Non-exclusive license to distribute, Open
Preprint
Action Guidance: Getting the Best of Sparse Rewards and Shaped Rewards for Real-time Strategy Games
arXiv (Cornell University)
05 Oct 2020
Abstract
Training agents using Reinforcement Learning in games with sparse rewards is
a challenging problem, since large amounts of exploration are required to
retrieve even the first reward. To tackle this problem, a common approach is to
use reward shaping to help exploration. However, an important drawback of
reward shaping is that agents sometimes learn to optimize the shaped reward
instead of the true objective. In this paper, we present a novel technique that
we call action guidance that successfully trains agents to eventually optimize
the true objective in games with sparse rewards while maintaining most of the
sample efficiency that comes with reward shaping. We evaluate our approach in a
simplified real-time strategy (RTS) game simulator called $\mu$RTS.
Metrics
9 Record Views
Details
- Title
- Action Guidance: Getting the Best of Sparse Rewards and Shaped Rewards for Real-time Strategy Games
- Creators
- Shengyi HuangSantiago Ontañón
- Publication Details
- arXiv (Cornell University)
- Resource Type
- Preprint
- Language
- English
- Academic Unit
- Computer Science (Computing)
- Other Identifier
- 991021869009804721