Modular Deep Reinforcement Learning
This session shows slides about deep maxpain algorithm we developed for robot navigation.
Policy Hyper-parameter Search
This session shows slides about policy-based reinforcement learning, including classic REINFORCE, PGPE, our propsed EPHE with K-elite, EPHE with adaptive baseline, EPHE with CMAES weighting, EPHE with REPS weighting methods, etc.
Smartphone Robot Project
This session shows how to construct a smartphone robot and achieve behaviors like standing-up, balancing, approaching, foraging and mating.