Leakage-Controlled and Information-Bounded Evaluation of Multi-Task Learning for Crumb–Rubber Modified Asphalt: What Twenty Mix Designs Can and Cannot Support

He Huang, Yingli Gao, Bin Tian, Zhuo Yang · 2026

Read Leakage-Controlled and Information-Bounded Evaluation of Multi-Task Learning for Crumb–Rubber Modified Asphalt: What Twenty Mix Designs Can and Cannot Support by He Huang, Yingli Gao, Bin Tian, Zhuo Yang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13567061.txt.

Description

He Huang and three colleagues published a methods study in two thousand twenty-six on machine learning for rubber-modified asphalt. Road binders mixed with tyre rubber are often modelled on small lab datasets. The same few mixes get tested at many temperatures, inflating the row count and inviting overconfident claims.

The team worked with two hundred twenty-one measurements from just twenty independent mixes. They compared sixteen models under strict, group-aware testing that holds out whole mixes. Before fitting anything, they proved the data itself caps how well one target property can ever be predicted. A locked hybrid model, mostly a robust classic branch, ranked first overall. The authors frame the paper as a lesson in honest evaluation.

Catalog record

CREATORHe Huang, Yingli Gao, Bin Tian, Zhuo Yang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13567061

Browse the full library · About the Tower of Babel