Osman Boyaci
I am a software research engineer at Intel. I work on machine learning, and particularly on graph neural networks — models that learn over data whose structure is a graph rather than a grid or a sequence.
- Now
- Software Research Engineer, Intel
- PhD
- Electrical and Computer Engineering, Texas A&M University, 2022
- MSc
- Computer Engineering, Istanbul Technical University, 2017
- BSc
- Computer Engineering and Electronics Engineering (double major), Istanbul Technical University, 2013
Background
My doctoral work, in Katherine Davis’s research group at Texas A&M, applied graph neural networks to the security of electric power systems: detecting and locating false data injection attacks in the measurements a grid operator depends on. A power network is already a graph, and the attacks worth worrying about are the ones built to survive the consistency checks a conventional state estimator performs — so the topology is where the remaining signal is. The dissertation was Graph Neural Network-Based Cybersecurity of Smart Grids.
Alongside the applications I have worked on graph learning methods themselves, and on problems outside power systems — including quantitative measurement of facial difference for craniofacial and plastic surgery.