Our summer courses are intensive 3-week programs (7+ hours/day) involving hands-on tutorials developed and taught by leading experts. Students are placed into TA-led groups of 10-20 students using the neuromatch algorithm, which matches students based on common interests, timezone and their preferred language. They receive personalized support as they work through hands-on tutorials and collaborate on course projects.

In order to capture all aspects of traditional summer schools, students will also be offered mentoring from professors in the field, professional development seminars, career panels and opportunities to virtually socialize with their fellow students.

All of our course materials are freely available to use in your own courses, teaching experiences, and other content as all of our materials have a CC-BY license.

Available Courses

We offer two courses: Computational Neuroscience and Deep Learning.

Computational Neuroscience

Starting with an optional 12 video series pre-course on basic neuroscience topics, a Python refresher course, and a mathematics refresher, students have everything they need to launch into the course most effectively.

The Neuromatch Academy computational neuroscience curriculum integrates cutting-edge advances in machine learning and causality reasearch with state-of-the-art modeling approaches in neuroscience. We teach in a code-first, hands-on Python tutorial-based format with teaching assistant (TA) support (interactive track only). But we don't just teach techniques: we emphasize interpretability and the process of modeling. Through group projects, participants get hands-on guided experience in how-to-model any observed phenomemon. We thus focus on policies, procedures, and processes: we teach a meta-science approach that translates to all of science! Group projects are supported by TAs and expert mentors in the field. This also results in professional networking opportunities that go beyond the project group.

The course starts with a intro to modeling, learning about what types of questions we can ask with given model types before allowing you to create your own. Then, you will move through the machine learning module, learning how to fit models to data, how to use generalized linear models and how to uncover underlying lower dimensional structure in data before finally learning how to use deep learning to build more complex models.

The dynamical system module teaches you to build more biologically plausible models. Here, coverage of topics like linear systems and dynamic networks allow you to build models that are based on bottom-up knowledge of the system we are modeling.

The stochastic processes module teaches you methods on how to get better insight through your measurement tools and how the state being measured can change over time (Hidden Dynamics). Then, you learn how to control these dynamic systems through optimal control and reinforcement learning methods.

Finally, you'll finish the course with an answer to one of the most important scientific questions: when can we determine if something is causally related vs. just correlated?

To see more, including the group projects, view the course schedule and course content here: Computational Neuroscience Course

Deep Learning

Our Deep Learning (DL) course grew out of the realization that there is a real need for teaching an ethically responsible hands-on TA-guided code-first DL curriculum that emphasizes how DL can be used to advance science and achieve better scientific insights. Science explains data, allows for generalization, and enables us to understand the goals of systems; it lets us understand causal inks and the mechanisms mediating these causes. DL approaches promise to help each of these aspects of scientific discovery because they are designed to handle the complexity of modern questions and datasets; and DL can thus complement the roles of statistics and scientists themselves.

To see more, including the group projects, view the course schedule and course content here: Deep Learning Course


Traditional summer schools can be expensive (~$5,000 for a three week course), but Neuromatch promises to be always affordable while still ensuring our teaching assistants get paid. To do this, we charge a much lower, regionally adjusted fee, and fee waivers for students that need them without any impact on their admission.

While Neuromatch is supported by generous donations from a variety of foundations and industry partners, we strive to make the program sustainable through tuition fees, which are used to pay our teaching assistants. We do not want fees to be a barrier for anybody, so these fees can be waived, but if you can pay even part of your fee, this is welcome. There are even opportunities to pay more than you fee if you would like to help subsidize fee waivers for other students with less financial means than yourself.

To estimate your course fees based on your region, use our COLA Calculator.

Previous Years

Neuromatch Academy 2021

Neuromatch Academy 2020