Interested in neuroscience, artificial intelligence, and prizes?

Then you are a great fit for a Neurodatathon! These events, sponsored by the Brown Institute for Brain Science and the Computation in Brain and Mind Initiative, provide instruction on state-of-the-art techniques in computation and Neuroscience and the opportunity to compete for prizes by analyzing special datasets that describe brain and behavior. Neurodatathons are open to all members of the Brown community, and those with backgrounds in Neuroscience, Computer science, Data science, and Applied Math are especially well suited to these events.

The 2017 Neurodatathon took place in mid-November and early-December. It began with workshops on programming, machine learning, deep learning, and Neuroscience, which ran from 11/16 Р11/17. Workshops were led by experts in one of these fields, who lectured and led programming tutorials to over 50 students with concentrations ranging from physics to neuroscience to computer science.

The competition itself ran from 11/28 – 12/2, and gave a total of 140 members of the Brown community the opportunity to compete for prizes by analyzing a one-of-a-kind dataset of neural and behavioral recordings. There were two competition tracks:

  1. CREATE, which rewarded students for creative analyses of the data.
  2. OPTIMIZE, where students competed to construct the most accurate computational model for decoding behavior from neural activity.

Winning submissions of the CREATE track included an analysis that found that different behaviors elicit different levels of periodicity in neural activity. Participants of the OPTIMIZE track found that pairing a technique called “regularization”, which can improve the generalization of a computational model, with a relatively small number of model parameters led to better decoding performance on this dataset than the kinds of “deep neural networks” that succeed in tasks like computer vision and speech recognition.

Stay tuned for more Neurodatathons next year! We are excited to expand our reach to more students interested in learning more about the interface of brain science and technology, so feel free to reach out with any questions you may have.