Explore memory augmented control networks in this informative lecture presented by Aravind Balakrishnan. Delve into the problem at hand and discover two key techniques: local observable area and value iteration networks. Gain insights into the differential computer architecture and its comparison with other methods. Examine practical applications through experiments in mazes and continuous control scenarios. Enhance your understanding of advanced control systems and their potential impact on artificial intelligence and robotics.
Memory Augmented Control Networks - Presenter: Aravind Balakrishnan
Overview
Syllabus
Introduction
The problem
Two techniques
Local observable area
Value iteration networks
Differential computer
Architecture
Comparison
Experiments
Mazes
Continuous Control
Conclusion
Taught by
Pascal Poupart
