Thor Martinsen, PhD

• Robustness and Vulnerability of Deep Learning Methods for Cyber Defense

  •  PI: CAPT Martinsen

  •  Co-PI: Prof. Kang

  •  Sponsor: Navy Research Program/ Navy Cyber Defense Operations Command

  •  Purpose: Investigation of the resiliency and vulnerability of Deep Learning methods

    applied in cybersecurity and network defense applications.

​

• Investigation of Cryptographic Properties of Generalized Boolean Functions

  • PI: CAPT Martinsen

  • Sponsor: Fulbright Foundation

  • Purpose: Investigation of generalized Boolean function security properties. Development of a

    set of functions suitable for use in cryptographic and communications protocols.

​

• Secure Machine Learning on Sensitive or Proprietary Data Sets

  •  PI: CAPT Martinsen

  • Sponsor: Omar N. Bradley Foundation, 2020

  • Purpose: Investigation of cryptographic techniques supporting secure Machine Learning on sensitive or proprietary data sets. This is a collaborative research project with a researcher from the Norwegian National Security Administration.

​

• An Improved Decimation Based Secure Pseudorandom Number Generator

  • PI: CAPT Martinsen, Co-PI: Prof. Stanica (MA)

  • Collaborators: Prof. Fouts (ECE)

  • Students, LT Vikram Kanth (ECE)

  • Sponsor: NPS Foundation (Seed)

  • Purpose: Improve security of cryptographic primitives & light-weight scream ciphers.

    This is a collaborative MA-CS-ECE project.

​