CLEAR Seminar with Dr. Jacqueline MacDonald Gibson, NC State University

Warning Icon This event is in the past.

When:
February 22, 2024
12:30 p.m. to 1:30 p.m.
Where:
Zoom link to be emailed to registrants
Event category: Seminar
Virtual
RSVP is closed.

Center for Leadership in Environmental Awareness and Research (CLEAR)
Virtual Seminar Bayesian Belief Networks for Predicting Exposure to Environmental Contaminants
Presenter: Jacqueline MacDonald Gibson, Ph.D., North Carolina State University
February 22, 2024 - 12:30 to 1:30 p.m. via Zoom

The Center for Leadership in Environmental Awareness and Research (CLEAR) is pleased to invite the campus community to a virtual seminar, "Bayesian Belief Networks for Predicting Exposure to Environmental Contaminants," on February 22, 2024 from 12:30 to 1:30 p.m. via Zoom. The seminar is free; registration is required. Registrants will receive the Zoom link via email.

Abstract

Traditionally, population exposure to environmental contaminants has been estimated using mechanistic models that account for the release, fate, and transport of contaminants in the environment or using geostatistical models that interpolate exposure concentrations using previously collected data.  With the rapid growth in the field of artificial intelligence, environmental engineers and scientists increasingly exploring the use of machine-learning tools for predicting the distribution of contaminants across space and time.  However, a limitation of these tools is that many of them are “black-box” approaches that mask relationships between factors that may influence contaminant exposure and the predicted exposure concentrations.  Bayesian networks, in contrast, are based on direct representations of influencing factors and the exposure variable of interest.  This presentation will provide background on Bayesian networks, including their origins and theoretical basis.  It will then present examples of machine-learned Bayesian networks for predicting the occurrence of contaminants in groundwater.  The accuracy of a machine-learned Bayesian network will be compared to that of a traditional mechanistic fate and transport model.

About Dr. Jacqueline MacDonald Gibson

Dr. MacDonald Gibson is the department head and professor of Civil, Construction and Environmental Engineering in the College of Engineering at North Carolina State University. Previously she was chair and professor of the Department of Environmental and Occupational Health in the School of Public Health at Indiana University, Bloomington. In addition, she was senior engineer at RAND Corporation, and associate director of the Water Science and Technology Board of the National Research Council. She received her Ph.D. in Engineering and Technology and in Civil and Environmental Engineering from Carnegie Mellon University. 

February 2024
SU M TU W TH F SA
28293031123
45678910
11121314151617
18192021222324
252627282912