Erica Shook '18 wants to flip the script on machine learning by using the technology to better understand the brain. Support from the W. Thomas Smith Scholarship will help her advance toward that goal in a post-graduate program at University College London.
Unique to Davidson, the Smith Scholarship is designed similarly to prestigious national postgraduate awards such as the Rhodes, Marshall, Luce and Fulbright scholarships. The award provides all expenses for two semesters of study at a major university outside of the United States.
Machine learning, a subfield of AI, uses algorithms inspired by the brain's own neural networks, Shook said. These algorithms are designed to help computers–hyper-literal machines–"learn" from experience and respond more effectively to ambiguous commands without explicit instruction from humans.
"I want to go in the opposite direction: How can machine learning help reveal the principles of neural computation?" she said.
Shook has honed theoretical and applied skills across disciplines, including neuroscience, math and computation, at Davidson.
Her first research mentors, in synthetic biology, were mathematician Laurie Heyer and genomicist Malcolm Campbell. Neuroscientist Julio Ramirez taught her stereotaxic neurosurgery, which makes use of a 3-D visualization control system, mathematician Donna Molinek introduced her to computational neuroscience through an independent study on the topic, and computer scientist Raghu Ramanujan helped her develop a machine-learning algorithm that predicts a rat's location by analyzing its patterns of neural activity.
Shook put her Davidson education on hold for a year when a summer lab internship at MIT turned into a year-long job offer. At MIT, she laid out an "in silico" computer brain that allowed her to measure each artificial neuron's response to any stimuli she chose to introduce to this "neural" network.
She presented her resulting auditory neuroscience research at the Association for Research in Otolaryngology in San Diego and will be an author on a forthcoming paper.
"Erica has a clear focus, a demonstrated track record, appropriate course work and abundant experience in her chosen field," said Campbell, Herman Brown Professor of Biology, director of the James G. Martin Genomics Program and current chair of the Fellowships & Scholarships Program. "Her employment by MIT before earning her bachelor's degree is astounding and a testament to her future potential in the field of computational neuroscience."
Shook's focus extends beyond her specific scientific interests to the broader goal of engaging and supporting women in science. She co-founded the student group FICSIT (Females in Computer Science and Information Technology), has taught and mentored other female scientists, and served as co-president of Davidson's student chapter of the Association of Women in Mathematics.
Davidson's Fellowships & Scholarships Program reflects the college's extraordinary commitment to supporting students' dreams–and plans–for lives of disproportionate impact beyond Davidson, said Jeanne-Marie Ryan, executive director of the Center for Career Development.
"Erica has a bright future in machine learning development and we're grateful to the Smith family for making this future possible," said Ryan.