Stephen J. Verzi Ph.D.

Stephen J. Verzi Ph.D.

Senior Member of Technical Staff

Contact Information

Stephen J. Verzi / 505-844-0063

Biography

Stephen J. Verzi earned his PhD from the University of New Mexico in computer science, with a focus in neural network design and analysis. At Sandia National Laboratories. His professional experience in algorithm development and computational modeling spans many application domains: artificial intelligence, machine learning, information retrieval and multilingual text analysis, neural function including the hippocampus and its substructures, system dynamics models of human behavior via the theory of planned behavior expressed using the mathematics of qualitative choice theory, individual-based and dynamical systems modeling, resilience metrics and their application in critical infrastructure, hospital evacuation and neural network dynamics, game-theoretic and agent-based models of adversarial behavior, and most recently exploring neural-inspired computation via phase-coding with application to image and video processing including anomaly detection as well as the application of adversarial reinforcement learning in grid resilience stability modeling. He helped develop and validate a population health effects model for the FDA/CTP currently used in policy-making there. He has over 45 peer-reviewed publications spanning the last 25 years.

Education

• Ph.D., Computer Science, University of New Mexico – 2003
• Master of Science, Computer Science, University of New Mexico – 1990
• Bachelor of Science, Computer Science, December 1987

U.S. Patents

• (App #12/364,753; Patent #US 8166051 B1), Bauer, T.L., Benz, Z.O., & Verzi, S.J. Computation of term dominance computation in text documents, Filed: 02/03/2009, Granted: 06/24/2012.
• (App #15/837,326; Pub. No.: US 2019/0180169 A1), Verzi, S.J., Vineyard, C.M., Miner, N.E., & Aimone, J.B. Optimization computation with spiking neurons, Filed: 12/11/2017, Published: 06/13/2019.
• (App #16/013,810) Devices and methods for increasing the speed or power efficiency of a computer when performing machine learning using spiking neural networks,,Vineyard, C.M., Aimone, J.B. Severa, W.M., & Verzi, S.J. Filed: 06/20/2018.
• Verzi, S.J., Vineyard, C.M., & Aimone, J.B. Anomaly detection with spiking neural networks (App #16/436,744), Filed 06/10/2019.

Google Scholar:

Stephen Verzi

Key Publications

2020

• Yoon, H., Melander, D., and Verzi, S. J., (2020), Permeability Prediction of Porous Media using Convolutional Neural Networks with physical properties [invited], AAAI 2020 Spring Symposium on “Combining Artificial Intelligence and Machine Learning with Physics Sciences”, March 23-25, 2020,

2019

• Severa, W., Vineyard, C.M., Dellana, R., Verzi, S.J. & Aimone, J.B. (2019). “Training deep neural networks for binary communication with the Whetstone method”. Nature Machine Intelligence, 1, pp. 86-94. doi: 10.1038/s42256-018-0015-y.

2018

• Galiardi, M.A., Verzi, S.J., Birch, G.C., Stubbs, J.J., Woo, B.L., & Kouhestani, C.G. (2018). “Physical Security Assessment Using Temporal Machine Learning”. Proceedings of the 2018 International Carnahan Conference on Security Technology (ICCST), pp. 1-5. doi: 10.1109/CCST.2018.8585705.
• Dawson, L. A., Verzi, S.J., Levin, D., Melander, D.J., Sorensen, A.H., Cauthen, K.R., Wilches Bernal, F., Berg, T.M., Lavrova, O., & Guttromson, R. “Integrated Cyber/Physical Grid Resiliency Modeling”. Technical Report, SAND2018-12934 669759. doi: 10.2172/1482777.