John Cunningham, a Columbia University researcher working at the intersection of statistics, machine learning and neuroscience, has received a 2015 Sloan Research Fellowship. An assistant professor of statistics, Cunningham is also affiliated with the Data Science Institute, the Grossman Center for the Statistics of Mind and the Zuckerman Institute. The Sloan Fellowship comes with a $50,000 award.

Trained in computer science, Cunningham as a graduate student at Stanford University went looking for the most challenging scientific problem that he could tackle with statistics and machine learning. He ended up in a neuroscience lab. “There’s nothing we use more and understand less at a computational level than our brains,” he said recently from his office overlooking northern Manhattan.

Neuroscientists had long recorded electrical pulses in the brain to understand its workings but rapid advances in computing power had now made it possible to analyze the system at scale. The problem that Cunningham chose to focus on, at least initially, was how the brain’s motor cortex controls precise movement through the firing and coordination of tens of millions of neurons. “The apparently simple act of raising a glass of water to my mouth is a profoundly difficult control problem,” he said.

Developing and applying algorithms to experimental data, Cunningham and his colleagues have begun to see the motor cortex in a new light. Unlike visual neurons, which fire predictably in response to pattern variations in the external world, motor neurons seemed to behave according to very different computational principles. The researchers described their work in a 2012 study in Nature, “Neural population dynamics during reaching.”

Mark Churchland, a neuroscience professor at Columbia who coauthored the paper and codirects the Grossman Center, compares the brain’s motor neurons to players in a band. “When the rhythms of all the players are summed over the whole band, a cascade of fluid and accurate motion results,” he told The Atlantic.

After earning his PhD from Stanford in 2009, Cunningham became a postdoctoral researcher in statistical machine learning at University of Cambridge. He came to Columbia in 2013, attracted by the chance to work with scientists at the frontiers of both statistical machine learning and neuroscience.  He continues to work on machine learning techniques to model and analyze complex neuronal signals to understand the brain’s enormous complexity.  

Other Columbia researchers to receive a 2015 Sloan Fellowship: mathematician Jennifer Horn; economist Suresh Naidu; geophysicist Tiffany Shaw and computational biologist Harris Wang.