I am a currently a Member of Technical Staff at MIT Lincoln Laboratory, where I research machine learning for RF signal processing. I received my PhD from the Department of Electrical and Computer Engineering at the University of Michigan in 2022, advised by Professor Laura Balzano.

My research lies in the intersection of machine learning, statistical signal processing, and optimization. I design and analyze new algorithms for problems with big messy data, using techniques grounded in principled modeling and supported by theoretical guarantees. My work spans a diverse range of applications, including signal processing, computer vision, medical imaging, banking, environmental sensing, and more. Some of my favorite research topics include

• Dimension reduction for high-dimensional data
• Optimization and efficient algorithms
• Online learning
• Data that are noisy, heterogeneous, incomplete, and/or corrupted.

Publications

Kyle Gilman, David Hong, Jeffrey Fessler & Laura Balzano. Streaming Heteroscedastic Probabilistic PCA with Missing Data, 2025. Transactions on Machine Learning Research. Paper

Kyle Gilman, Sam Burer, & Laura Balzano. A Semidefinite Relaxation for Sums of Heterogeneous Quadratic Forms on the Stiefel Manifold, 2025. SIAM Journal on Matrix Analysis and Applications. Paper

Kyle Gilman, Davoud Ataee Tarzanagh, & Laura Balzano. Grassmannian Optimization for Online Tensor Completion and Tracking with the t-SVD, 2022. IEEE Transactions On Signal Processing. 70 pp. 2152-2167. Paper

David Hong, Kyle Gilman, Laura Balzano, & Jeffrey A. Fessler. HePPCAT: Probabilistic PCA for Data with Heteroscedastic Noise, 2021. IEEE Transactions on Signal Processing, Vol. 69, 2021. Paper

Kyle Gilman and Laura Balzano. Online Tensor Completion and Free Submodule Tracking with the t-SVD, 2020. ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Paper

Kyle Gilman and Laura Balzano. Panoramic Video Separation with Online Grassmannian Robust Subspace Estimation, 2019. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Paper