How can we address the current limitations of reinforcement learning: sample efficiency, scaling, and generalization? @yfletberliac gives a comprehensive overview of one potential path forward: hierarchical reinforcement learning. https://t.co/QHMqpYNtmx
— The Gradient (@gradientpub) March 9, 2019
from Twitter https://twitter.com/gradientpub
March 09, 2019 at 12:24PM
via IFTTT
No comments:
Post a Comment