- Research Program Mentor
PhD Doctor of Philosophy candidate
machine learning (especially reinforcement learning and meta-learning); robotics, Computer Science, reinforcement learning, AI/ML, deep learning, math
Project ideas are meant to help inspire student thinking about their own project. Students are in the driver seat of their research and are free to use any or none of the ideas shared by their mentors.
Exploring practical algorithmic choices in deep reinforcement learning
Despite recent successes in deep reinforcement learning, many techniques commonly used in the field (e.g. target networks, clipping, maximum entropy) are poorly understood. The project would experimentally probe the necessity of these tricks, and potentially explore other improvements.