We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
ADDITIONAL OBSERVATIONS ON THE SITE RECOGNITION CHALLENGE.
- Authors
Kok-Kwang Phoon; Jianye Ching
- Abstract
One distinctive feature of geotechnical engineering is site uniqueness or site-specificity. However, there is no data-driven method to quantify site uniqueness. The corollary is that it is not possible to identify "similar" sites from big indirect data (BID) automatically and no method to combine sparse site-specific data with big indirect data to produce a quasi-site-specific model that is less biased compared to a generic model and less imprecise compared to a site-specific model. This "site recognition" challenge is difficult because site-specific data is MUSIC-X (Multivariate, Uncertain and Unique, Sparse, Incomplete, and potentially Corrupted with "X" denoting the spatial/temporal dimension). This paper presents the application of four data-driven methods (hybridization, hierarchical Bayesian model, record similarity method, site similarity method) to construct a quasi-site-specific transformation model between the undrained shear strength and normalized cone tip resistance. The similarity methods are "explainable", because a list of "similar" sites can be generated explicitly for inspection by the engineer. The effect of extrapolating the quasi-site-specific model beyond the range of the training dataset is also studied by comparing the performance of these models under routine validation (validation dataset is contained within the training dataset) and under external validation (validation dataset lies outside the training dataset). The hierarchical Bayesian model appears to be the best performing method thus far, but it suffers from a lack of "explainability". More research is needed to: (1) ascertain the number and/or type of soil properties needed to identify "similar" sites more robustly in the sense of producing more clustered results in existing soil classification charts (e.g., Casagrande plasticity chart, Robertson CPT-based soil behavior type classification system) and/or producing the most accurate quasi-sitespecific model, (2) understand the bias and precision of making inferences beyond the range of the training dataset, and (3) clarify the trade-off between explainability and inference (bias and precision).
- Subjects
ENGINEERING inspection; SOIL classification; SHEAR strength; GEOTECHNICAL engineering
- Publication
Journal of GeoEngineering, 2022, Vol 17, Issue 4, p231
- ISSN
1990-8326
- Publication type
Article
- DOI
10.6310/jog.202212_17(4).6