Xuezhou (Jack) Zhang

Associate Research Scholar

ECE Department and Center for Statistics and Machine Learning

Princeton University 

Reinforcement Learning (RL) is a form of machine learning where an autonomous agent learns how to behave in an unknown environment through active interactions. Unlike traditional supervised learning, the agent has control over its environment and must gather feedback to make predictions. Despite its potential, RL has faced challenges in achieving success in real-life applications. In this talk, I will describe two main obstacles in RL and offer solutions.

The first challenge is data efficiency. Deep RL algorithms are often highly data-inefficient, requiring large amounts of interactions to find a good policy. I will discuss how representation learning can provide a fundamental solution to this problem, both in theory and in practice. In particular, I will talk about how a low-dimensional representation can be learned from high-dimensional observations in an online fashion, and how such representations can facilitate efficient learning in current and future tasks.

The second challenge is security and robustness. To learn from raw interactions, an RL agent must be resilient against noisy data, distribution shifts, and even adversarial attacks. I will talk about how wisdom from traditional robust statistics can help us transform deep RL algorithms to achieve robust learning even against the strongest adversaries.
Finally, I will conclude my talk with future work directions that make AI interact more efficiently with one another and with human users.

Xuezhou (Jack) Zhang

Associate Research Scholar

ECE Department and Center for Statistics and Machine Learning

Princeton University 

Reinforcement Learning (RL) is a form of machine learning where an autonomous agent learns how to behave in an unknown environment through active interactions. Unlike traditional supervised learning, the agent has control over its environment and must gather feedback to make predictions. Despite its potential, RL has faced challenges in achieving success in real-life applications. In this talk, I will describe two main obstacles in RL and offer solutions.

The first challenge is data efficiency. Deep RL algorithms are often highly data-inefficient, requiring large amounts of interactions to find a good policy. I will discuss how representation learning can provide a fundamental solution to this problem, both in theory and in practice. In particular, I will talk about how a low-dimensional representation can be learned from high-dimensional observations in an online fashion, and how such representations can facilitate efficient learning in current and future tasks.

The second challenge is security and robustness. To learn from raw interactions, an RL agent must be resilient against noisy data, distribution shifts, and even adversarial attacks. I will talk about how wisdom from traditional robust statistics can help us transform deep RL algorithms to achieve robust learning even against the strongest adversaries.
Finally, I will conclude my talk with future work directions that make AI interact more efficiently with one another and with human users.