Comparative machine learning models for predicting the compressive strength of ultra-high-performance concrete
Nguyen Thi Thu Nga, Nguyen Anh Tuan, Nguyen Tuan Tu, Huyen Thi Dang · 2026
Read Comparative machine learning models for predicting the compressive strength of ultra-high-performance concrete by Nguyen Thi Thu Nga, Nguyen Anh Tuan, Nguyen Tuan Tu, Huyen Thi Dang (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/plos-pone-0357153.txt.
Description
Nguyen Thi Thu Nga leads a 2026 PLOS ONE comparison of machine learning models for predicting the strength of ultra-high-performance concrete. The material is exceptionally strong and durable, but its compressive strength depends on a complex mix design. Lab testing is slow, so data-driven prediction is attractive.
Five advanced algorithms were trained on an experimental database. CatBoost performed best, explaining over ninety percent of strength variation with modest average errors. Yet for the strongest mixes, at or above a high threshold, errors grew and the model tended to underpredict, because such samples were few. Explanation tools named the mix variables that mattered most. The model suits early mix screening within well-covered strength ranges.
Catalog record
| CREATOR | Nguyen Thi Thu Nga, Nguyen Anh Tuan, Nguyen Tuan Tu, Huyen Thi Dang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59403 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |