Class Goals

The goal of the course is to introduce you to introduce you to the basic mathematics and algorithms needed for research on brain machine interfaces. The course will also provide an overview of the current research in this area.  this is an exciting new field and results are coming quickly and consequently we'll be reading papers and using data that is at the state of the art.

A partial list of topics to be covered:

Basic neuroscience.

Statistical models of neural firing.

Population coding.

Neural decoding.

Spike "sorting".
Basic probability theory.
Bayesian inference.
Linear regression.
Principal Component Analysis (PCA).
Mixture Models and Expectation-Maximization (EM).
Kalman filtering.

Particle filtering and Monte Carlo sampling.
User interfaces for the disabled.
Prosthetics (vision, limbs, auditory)

Our goals are teach you some basic tools that can be used in many disciplines (machine learning, computer vision, graphics, natural language understanding, ...), expose you to the current state of the art in BCI, and stimulate you to thinking about the future (since the future of this field is wide open).

The course is organized as a graduate seminar and one of the goals is to get you familiar with reading papers critically and efficiently. 

Class Structure

The class meets for three hours on Wednesdays.   A typical class might have a lecture style presentation of some key material in the first half followed by discussions of readings or experiments in the second half.

This is a "learn-by-doing" sort of class. There will be assignments throughout that give you hands-on experience with the math and algorithms applied to real data.

Readings will be from current research papers and you will have to hand in a review of one paper every week.

Finally, I want people to participate in the class discussions and consequently some of your grade will be for participation.

Return to CS295-7 home page.

Return to Michael Black's home page.