معرفی
James Murphy is an Associate Professor of Mathematics at Tufts University, with affiliate appointments in Computer Science and Electrical and Computer Engineering. His research focuses on the mathematics of artificial intelligence, statistical and machine learning, high-dimensional probability, and their applications to image/signal processing, molecular dynamics, and computational biology. He holds grants from the NSF, NGA, Camille & Henry Dreyfus Foundation, and others. His work integrates theoretical mathematics with computational tools, emphasizing open-source software development.
Education: PhD in Mathematics (2015) and MS (2013) from University of Maryland, BS in Mathematics (2011) from University of Chicago.
Research interests include machine learning theory (e.g., spectral clustering, unsupervised learning), harmonic analysis, graph theory, and applications to hyperspectral imaging, medical data analysis, and remote sensing. Notable projects involve Wasserstein-based methods for data clustering, diffusion geometry for image processing, and algorithms for anomaly detection in high-dimensional datasets.
His research group, HILDA, collaborates on projects like GLIDE (biological network prediction), Fermat distances for clustering, and entropy-regularized optimal transport. He teaches courses in functional analysis and has contributed to open-access educational materials for calculus and algebra.
