Recent advancements in machine learning, while powerful, are often burdened by significant computational and memory requirements, limiting their deployment in resource-constrained settings. Hyperdimensional Computing (HDC) emerges as an alternative with its simplicity, lightweight operations, and robustness to errors in high dimensional space. By encoding data into high-dimensional vectors and performing efficient algebraic computations, HDC opens a new avenue as an efficient learning paradigm.
In this talk, I will introduce the fundamentals of HDC and briefly review existing research across its various stages. I will then present PIONEER, a novel approach that uses learned projection vectors to optimize the encoding process. By leveraging neural networks to learn these vectors, PIONEER allows HDC to achieve high accuracy while preserving efficiency. Finally, I will discuss two applications of HDC: a cost-effective, noise-resilient pressure mat system for human activity recognition and the detection of epileptic seizures from surface EEG.
Fatemeh Asgarinejad is an Assistant Professor of Teaching in the Electrical and Computer Engineering Department at the University of California, Riverside. Prior to UCR, she earned her Ph.D. in Electrical and Computer Engineering from the University of California, San Diego. She is the recipient of the 2025 Excellence in Teaching Award from UC San Diego’s Computer Science and Engineering Department and 2025 Barbara J. and Paul D. Saltman Excellent Teaching Award. Her Ph.D. research focused on the synergy between brain-inspired Hyperdimensional Computing and Machine Learning. She is now advancing innovations in Electrical and Computer Engineering education and exploring applications of machine learning.