November 11, 2015
Report

UQ Methods for HPDA and Cybersecurity Models, Data, and Use Cases

Abstract

This report describes our initial research to quantify uncertainties in the identification and characterization of possible attack states in a network. As a result, we should be able to estimate the current state in which the network is operating, based on a wide variety of network data, and attach a defensible measure of confidence to these state estimates. The output of this research will be new uncertainty quantification (UQ) methods to help develop a process for model development and apply UQ to characterize attacks/adversaries, create an understanding of the degree to which methods scale to "big" data, and offer methods for addressing model approaches with regard to validation and accuracy.

Revised: June 5, 2018 | Published: November 11, 2015

Citation

Engel D.W., K.D. Jarman, Z. Xu, B. Zheng, A.M. Tartakovsky, X. Yang, and R. Tipireddy, et al. 2015. UQ Methods for HPDA and Cybersecurity Models, Data, and Use Cases Richland, WA: Pacific Northwest National Laboratory.