Karamatou Yacoubou Djima is a Beninese applied mathematician and an Assistant Professor of Mathematics at Amherst College. Before moving to Amherst, she spent a year at Swarthmore College as a visiting postdoctoral fellow.
She received both her Ph.D. (2015) and MSc (2011) in Applied Mathematics & Statistics and Scientific Computing from the University of Maryland in College Park, under the supervision of Wojciech Czaja. She studied mathematics and engineering sciences at CUNY-CSI, where she enjoyed her first research experiences in mathematics (applied dynamical systems) under the supervision of Andrew Poje and physics (supersolids) under the supervision of Anatoly Kuklov. Before living in the US, Karamatou spent her life between Ivory Coast, Togo, and Benin.
Karamatou’s current research interests lie at the intersection of applied harmonic analysis and machine learning. Her goal is to represent data (for example, images, human voice, or social media trends) using simple building blocks that unveil hidden structures in the data, extract relevant information, and reduce the cost of computational algorithms. Her past and ongoing projects include novel spectral graph methods, early diagnosis of autism spectrum disorder using features present in placenta images, detection of age-related macular degeneration via identification of anomalies in retinal images, and motion detection in animated images for Pixar.