BC3 Neural Population Dynamics during Learning


What is the neural basis of perception, action, and cognition? Now that we can record the activity of populations of neurons acting together, new answers are emerging. To interpret population recordings, we must draw on tools from machine learning. In this course, the chief application of these ideas is in brain-computer interfaces.
Lecture 1 will introduce sensory-motor neurophysiology from a conventional vantage point. The nonhuman primate model has provided the richest opportunity to relate cognition to neural activity.
Lecture 2 will introduce general concepts in machine learning, focusing on the objectives of each technique, the motivating intuitions, and examples of how they yield new insight in neuroscience.
Lecture 3 examines brain-machine interfaces, from both clinical and basic-science perspectives.
Lecture 4 considers applications of machine learning to neurophysiology more broadly, focusing on decision-making, attention, and the future of neural prosthetics.


Students will:
Learn organizing principles of brain function, in particular the cerebral cortex
Learn concepts from machine learning, and how they apply to neuroscience data
Learn principles, techniques, and advances in the field of neural prosthetics.


“Dimensionality Reduction for Large-Scale Neural Recordings”
Yu and Cunningham

“Brain-computer interfaces for dissecting cognitive processes underlying sensorimotor control”
Golub, Chase, Batista and Yu

“Neural Constraints on Learning”
Sadtler et al.

Course location


Course requirements


Instructor information.

Aaron Batista


University of Pittsburgh


Aaron Batista is an Associate Professor in the Department of Bioengineering at the University of Pittsburgh. His undergraduate education was in philosophy and computer science at the University of Pennsylvania. He earned a PhD in Computation and Neural Systems at Caltech. He conducted his postdoctoral research at Stanford University with Profs. Krishna Shenoy and Bill Newsome.