DATA8016 - Quantitative Neuroscience and AI: Modeling, Inference, and Shared Principles
Course Instructor
Professor Andrew LUO
Assistant Professor
HKU Musketeers Foundation Institute of Data Science and
Department of Psychology
Professor Andrew Luo is an Assistant Professor at the Institute of Data Science (IDS) and the Department of Psychology at The University of Hong Kong. Professor Luo obtained his PhD from Carnegie Mellon University in the joint program in Neural Computation & Machine Learning, advised by Prof. Michael J. Tarr and Prof. Leila Wehbe. Previously he obtained his Bachelor of Science from MIT in Computer Science.
Professor Luo’s research interests lie at the intersection of computer vision, human visual representations, scene learning, and generative models. His research focus is on developing machine learning models that can understand and perceive the world like humans, bridging the gap between cognitive science and artificial intelligence.
Course Description
This doctoral level class examines the interdisciplinary nexus of neuroscience and artificial intelligence (NeuroAI). The course provides doctoral students with core theoretical frameworks and methodological skills essential for conducting original research in this field. It is designed for students specializing in either machine learning or neuroscience, establishing a foundational basis for NeuroAI research. The curriculum critically evaluates two interconnected themes: the application of contemporary artificial intelligence architectures as models of neural function, and the potential utility of neuroscientific principles for advancing artificial intelligence.
The course integrates two primary components: (1) Critical Analysis: Examination of current methodologies and theoretical debates in computational neuroscience through lectures, assigned readings, and structured discussion. (2) Empirical Research: A substantial, project-based component involving computational assignments and an independent research project. This project is designed to enable students to formulate empirically grounded perspectives and has the potential to yield publishable results. Course content focuses on the comparative analysis of representational strategies in biological neural systems (across species) and artificial intelligence models.
Core domains include: sensory processing, motor control, language, and higher cognition. Students will learn and critically assess state-of-the-art techniques for employing AI models as hypotheses of brain function, including their strengths and limitations. The seminar will also address generative mechanisms in both biological and artificial systems and explore the potential for their integration.
Prerequisites
Proficiency in intermediate statistics or machine learning is required. Students must possess working knowledge of deep learning fundamentals, linear algebra, and Python programming. Experience with array-based computation libraries (e.g., PyTorch, NumPy, TensorFlow) is essential.
Prior coursework in neuroscience is advantageous but not mandatory. Students without neuroscience background will receive introductory materials at the semester outset. Prior coursework in neuroscience is advantageous but not mandatory. Students without neuroscience background will receive introductory materials at the semester outset. Students with prior trainings or experiences should consult the instructor before registration.
The course includes structured support for scholars transitioning from non-neuroscience disciplines. Prospective students uncertain about prerequisite fulfillment should reach out to the instructors!