Hierarchical Reinforcement Learning
Doina discusses the potential for applying linear programming methodologies to reinforcement learning, particularly in scenarios where preferences are not clearly defined. She highlights the challenge of evolving preferences over time and the importance of discovering abstractions through frameworks like options, which allow agents to represent temporally extended actions effectively. The conversation opens up intriguing possibilities for practical applications and future research in the field.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Hierarchical and Continual RL with Doina Precup - #567
Related Questions