Clavius Distinguished Lecture with Bruce S. Kristal, Ph.D.

Bruce S. Kristal, Ph.D., will present “Modeling at a Different Level to Optimize Data Use: Accurately Predicting the Outcome of the Fusion of Mathematical Models.”

Resolving technological gaps in the optimal handling and leveraging of data resources will facilitate future progress across many fields. Often, it is advantageous to combine multiple mathematical models for robustness or power, but realizing these potential gains cannot be guaranteed. This fundamental technological gap limits our ability to leverage data across essentially any domain in which more than one potentially useful mathematical model exists. In this lecture, Kristal will discuss the development of the DIRAC framework, an approach that complements, not replaces, standard modeling approaches by accurately predicting the outcome of both score and rank-based fusions. DIRAC’s core implementation is inherently domain- and distribution-independent and has predicted utility across far-ranging areas, such as clinical biomarker development/personalized medicine, clinical trial enrollment, insurance pricing, portfolio management, and sensor optimization.