OpenAI News·· 2017-03-24AI 评分47
OpenAI 发现进化策略可替代强化学习
Evolution strategies as a scalable alternative to reinforcement learning
AI 导读
OpenAI 研究发现，进化策略（ES）在 Atari 和 MuJoCo 等现代基准测试中性能可与标准强化学习（RL）相媲美。该技术克服了 RL 的诸多不便，为大规模优化提供了可扩展的替代方案。
原文
We’ve discovered that evolution strategies (ES), an optimization technique that’s been known for decades, rivals the performance of standard reinforcement learning (RL) techniques on modern RL benchmarks (e.g. Atari/MuJoCo), while overcoming many of RL’s inconveniences.
来源：OpenAI News · openai.com