Multi-Agent Systems

Pushing the limits of scaling

Multi-agent systems is where we live our lives. Learning how to model and interact with 1000s of agents is one of the end-games of learning and control. Current methods do not scale upwards of 10s of agents! I entered this space from the lens of mobile sensor networks and swarm-v-swarm combat! Just getting started.

‍

Open areas of study

• multi-agent model-based RL,
• using deep learning to extend control policies to 1000s of agents,
• anti-swarm combat,
• applications of drones.