Beyond Aleatory---A Taxonomy of Uncertainty for Reduction and Consequence
This event is open to:
ICME Seminar: "Beyond Aleatory---A Taxonomy of Uncertainty for Reduction and Consequence" - Monday, June 6, 2022 - 4:30-5:30pm PDT.
Abstract: Researchers in uncertainty quantification (UQ) often dichotomize uncertainty as *epistemic* or *aleatory*---reducible or irreducible. While fundamental, this dichotomy is insufficient to fully describe uncertainties of practical interest. Namely---the aleatory/epistemic dichotomy 1. does not intrinsically suggest *how* to reduce uncertainty, and 2. does not consider the *consequences* of uncertainty. Concepts from industrial statistics and statistics education address these limitations: The notions of *chance and assignable cause* serve as an actionable framework for reducing uncertainty, while *real and erroneous source* articulate the consequences of uncertainty. In this talk I will detail these cause-source axes, and will provide examples of new research questions that can be posed using this framework.
Speaker Bio: Zachary del Rosario is an Assistant Professor of Engineering and Applied Statistics at Olin College. His goal is to help scientists and engineers reason under uncertainty. Zach uses a toolkit from data science and uncertainty quantification to address a diverse set of problems, including reliable aircraft design and AI-assisted discovery of novel materials.