Managing uncertain ground truth using Bayesian machine learning
Journal
Proceedings of the 29th European Safety and Reliability Conference, ESREL 2019
Pages
9-21
Date Issued
2020
Author(s)
Abstract
Two disparate trends can be easily discerned. One, digital technologies are evolving by leaps and bounds. The volume, variety, and velocity of data can only increase. A geotechnical engineer will soon be asking what to do with this deluge of data. This is a fundamental shift from an existing environment that is data poor. Second, engineered systems are increasing in scale, complexity, interconnectivity, among others and the emerging resilience engineering paradigm in response to this challenge is to design for both expected and unexpected conditions. There is no precedent if a condition is truly unexpected. Although the geotechnical engineering profession has been successful in managing uncertain ground truth with very limited data, this practice is unlikely to meet these challenges and to exploit new opportunities. This paper discusses the application of Bayesian machine learning to characterize site effects and to estimate soil/soil properties under a set of general constraints abbreviated as MUSIC-X (Multivariate, Uncertain and Unique, Sparse, Incomplete, and potentially Corrupted data with variations in space/time). More research beyond this modest start is needed to understand how to exploit existing databases along this data-driven pathway to support decision making.
SDGs
Type
conference paper