John P. Cunningham

Research interests:

1. Developing machine learning methods to analyze motor cortical processing
   The motor system, and brain systems in general, are complicated and poorly understood. Using machine learning and optimization (as well as information theory and signal processing), I study ways to infer meaningful information from neural spike data. This has benefits both for neural prosthetic systems and for neuroscientific studies in general. Example.
2. Developing fast computational methods for machine learning
   Many machine learning and optimization methods have appealing convergence properties and theoretical guarantees, but realistic run-time and memory tractability (particularly in the face of huge data sets) is still a challenge. I employ large scale optimization techniques and machine learning approximations to enable computational practicality. Example.
3. Algorithm design for neural prosthetic systems
   Neural prosthetic systems seek to restore motor function and communication ability to disabled people. This effort relies on fast, adaptable algorithms that can optimize the performance of a prosthetic system, given a noisy neural signal. This work involves a mixture of machine learning, optimization, and information theory. Example.
4. Experimental Electrophysiology
   All of the above methods prove their merit in their application to real neural data. As part of the Neural Prosthetic Systems Laboratory, I use implantable array technology to record simultaneously from tens to hundreds of motor cortical neurons, and I design experimental paradigms to scientifically validate the methods above. Example.

Education:

• Stanford University, Ph.D. Candidate Electrical Engineering (tentatively 2009)
• Stanford University, M.S. Electrical Engineering 2006
• Dartmouth College, A.B. Computer Science 2002

Fellowships and Awards:

• Michael Flynn Stanford Graduate Fellowship 2004-present
• Earned 8th place (of 160) in the Stanford E.E. Ph.D. Qualifying Exams 2006
• Rufus Choate Scholar, Dartmouth College 2002
• Phi Beta Kappa 2002-present

Teaching:

• TA in Stanford Electrical Engineering classes, beginning Fall 2008
• Math and Science Tutor, Foundation for a College Education 2007-Present
• Stanford Electrical Engineering Quals Tutor 2007

Work:

• Technical consultant for private companies and investors, 2006-present
• Morgan Stanley, Investment Banking Analyst, Technology Mergers and Acquisitions, 2002-2004
• Cisco Systems, Customer Support Engineer and Programmer, 2000-2002
• Edgartown Police Department, Special Officer, 1999

Talks:

• "Fast Gaussian Process Methods for Point Process Intensity Estimation," at the 25th International Conference on Machine Learning, July 7, 2008
• "Practical Optimization Tricks and Tips," at the Gatsby Computational Neuroscience Unit, University College London, London, UK, June 19, 2008
• "Engineering Challenges in Neural Prosthetic Systems," at the Stanford Bioengineering Forum, February 26, 2008
• "Neural Basis of Reach Preparation and Neural Communication Prostheses," at the Neukom Institute for Computational Science, Dartmouth College, February 11, 2008
• "Engineering Challenges in Neural Prosthetic Systems," at the Thayer School of Engineering, Dartmouth College, February 8, 2008
• "Inferring Neural Firing Rates from Spike Trains using Gaussian Processes," Spotlight Presentation at Neural Information Processing Systems 20 (NIPS 20), Vancouver, December 5, 2007
• Research Talk, Gatsby Computational Neuroscience Unit, University College London, London, UK, November 27, 2007