ICLR 2020: Yann LeCun and Energy-Based Models
Machine Learning Street Talk (MLST)
00:00
Reinforcement Learning vs. Model Predictive Control
This chapter explores the differences between traditional reinforcement learning and model predictive control approaches, emphasizing the use of energy functions to evaluate car trajectories. It discusses the complexities of planning and following road rules in autonomous navigation, while highlighting the philosophical implications of various learning models.
Transcript
Play full episode
