Abstract: This paper presents a novel Bayesian machine-learning approach for unsupervised and simultaneous soil stratigraphic interpretation of cone penetration test (CPT) soundings at the site scale. The proposed approach interprets ...