in Verification, Validation, and Uncertainty Quantification
The Future Of Risk Assessment
Clarity for those organizations that need it most.
Enhancing Risk Models
The Short Course
Alexandria Validation Consulting is owned and operated by Michael Scott Balch, PhD, an aerospace engineer who has specialized in uncertainty quantification for over ten years. The company is open to clients in any field who need help quantifying uncertainty in empirically rigorous terms. Uncertainty is, after all, a universal human experience.
In his doctoral work, Dr. Balch pioneered the application of random sets as a method for obtaining probability bounds in atmospheric uncertainty analysis, in support of NASA's Mars program. He also provided a vigorous defense for adhering to frequent standards of belief in analyses done in support of an engineering safety case. This defense illustrated how the practical constraints of a problem in uncertainty quantification drive the choice of epistemology, which in turn drives the choice of method. This connected triple, problem-epistemology-method, is central to the pedagogy of the short courses taught at Alexandria Validation Consulting, LLC.
In subsequent posts, Dr. Balch has produced new methods for engineers and scientists working in uncertainty quantification. At Applied Biomathematics, as part of a project for the NIH, Dr. Balch laid down the foundations for the theory of confidence structures, a frequentist alternative to Bayesian parameter inference. At Wright-Patterson Air Force Base, Dr. Balch developed a hybrid Bayesian-significance testing approach for assessing model form hypotheses, in support of AFRL's hypersonics validation efforts. Most recently, Dr. Balch has introduced a judiciously tailored random-to-fuzzy transformation as a fix for probability dilution in satellite conjunction analysis. While most practitioners spend decades using the same two or three statistical tricks to solve well-established problems, Dr. Balch has spent the past ten years aggressively seeking out the most difficult problems in uncertainty quantification and solving them using either new or cutting-edge techniques.
V&V&UQ are the mathematics that analysts like us do to make sure that model-informed decisions don't blow up in our clients' faces. We make sure that quantitative models are executed correctly (verification). We make sure that quantitative models are adequately faithful to reality (validation). And we make sure that uncertainties in the inputs to a quantitative model are accounted for correctly (uncertainty quantification). In short, V&V&UQ is the analysis that you do to make sure that your nominal risk analysis is as accurate as possible.
Far too often in engineering and finance today, the nominal risk analysis is far from acceptable. In finance, look no further than the 2008 crash. In engineering, look no further than the latest (usually preventable and predictable) space launch system failure. While a few analysts are in denial, for most, it is an open secret that traditional risk analysis methods are thoroughly inadequate to modern needs.
Alexandria Validation Consulting, LLC exists to completely fix that problem.
David is responsible for gathering information on potential fields of application for our V&V&UQ methodology. He provides many clients with their first introduction to Alexandria Validation Consulting, LLC. With several years of experience in management, sales, and promotion, David has demonstrated a consistent ability to adapt to new problems and successfully guide organizations through transition and expansion.
Dr. Balch is the principal instructor and chief consultant at Alexandria Validation Consulting, LLC. At present, all technical services are provided by Dr. Balch personally. With ten years of experience in verification, validation, and uncertainty quantification; Dr. Balch has established a reputation for delivering practical solutions to the most challenging problems.