Search

Subjective judgements from experts provide essential information when assessing and modelling threats in respect to cyber-physical systems. For example, the vulnerability of individual system components can be described using multiple factors, such as complexity, technological maturity, and the availability of tools to aid an attack. Such information is useful for determining attack risk, but much of it is challenging to acquire automatically and instead must be collected through expert assessments. However, most experts inherently carry some degree of uncertainty in their assessments. For example, it is impossible to be certain precisely how many tools are available to aid an attack. Traditional methods of capturing subjective judgements through choices such as \emph{high}, \emph{medium} or \emph{low} do not enable experts to quantify their uncertainty. However, it is important to measure the range of uncertainty surrounding responses in order to appropriately inform system vulnerability analysis. We use a recently introduced interval-valued response-format to capture uncertainty in experts' judgements and employ inferential statistical approaches to analyse the data. We identify key attributes that contribute to hop vulnerability in cyber-systems and demonstrate the value of capturing the uncertainty around these attributes. We find that this uncertainty is not only predictive of uncertainty in the overall vulnerability of a given system component, but also significantly informs ratings of overall component vulnerability itself. We propose that these methods and associated insights can be employed in real world situations, including vulnerability assessments of cyber-physical systems, which are becoming increasingly complex and integrated into society, making them particularly susceptible to uncertainty in assessment., Comment: International Conference on Critical Information Infrastructures Security (CRITIS) 2019

Technological advances in genotyping have given rise to hypothesis-based association studies of increasing scope. As a result, the scientific hypotheses addressed by these studies have become more complex and more difficult to address using existing analytic methodologies. Obstacles to analysis include inference in the face of multiple comparisons, complications arising from correlations among the SNPs (single nucleotide polymorphisms), choice of their genetic parametrization and missing data. In this paper we present an efficient Bayesian model search strategy that searches over the space of genetic markers and their genetic parametrization. The resulting method for Multilevel Inference of SNP Associations, MISA, allows computation of multilevel posterior probabilities and Bayes factors at the global, gene and SNP level, with the prior distribution on SNP inclusion in the model providing an intrinsic multiplicity correction. We use simulated data sets to characterize MISA's statistical power, and show that MISA has higher power to detect association than standard procedures. Using data from the North Carolina Ovarian Cancer Study (NCOCS), MISA identifies variants that were not identified by standard methods and have been externally ``validated'' in independent studies. We examine sensitivity of the NCOCS results to prior choice and method for imputing missing data. MISA is available in an R package on CRAN., Comment: Published in at http://dx.doi.org/10.1214/09-AOAS322 the Annals of Applied Statistics (http://www.imstat.org/aoas/) by the Institute of Mathematical Statistics (http://www.imstat.org)

This study provides information on obstacles facing homeless youth in school. Research occurred in four diverse New England cities. Researchers collected detailed case histories on youth age 10-15 years who were currently homeless or who had recently been homeless. Data came from staff of local youth agencies, government officials, and youths attending a 90-minute focus group discussion on their personal experiences and observations. Focus group participants also completed a five-page questionnaire. Most participants were female and white, though some were African American and Hispanic American. All had attended public school at some time, and over half were unemployed. Most were living in federally funded transitional living programs, state-run independent living or parenting teen programs, or foster care. The study looks at the impact of the Stewart B. McKinney Homeless Assistance Act. Results indicate that the academic experiences of responding youth were overwhelmingly negative. Common problems included transportation, inability to catch up after missed classes, clashes with authority figures, and perceived lack of respect from school personnel. Some respondents were depressed and anxious. For youth living outside of McKinney-funded districts, the Act had little impact on educational changes or experiences. Remedies to this situation are discussed. (Contains 22 references.) (SM)