Translating dietary standards into healthy meals with few-ingredient substitutions
Trevor Chan, Ilias Tagkopoulos · 2026
Read Translating dietary standards into healthy meals with few-ingredient substitutions by Trevor Chan, Ilias Tagkopoulos (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-13218539.txt.
Description
Trevor Chan and Ilias Tagkopoulos present a 2026 computational framework that turns dietary guidelines into meals with tiny changes. Personalized diet tools often optimize a single goal or demand unrealistic overhauls, so their advice goes unused. The goal here was the opposite: start from meals people eat and swap as little as possible while hitting nutrition targets.
Working from a national intake dataset covering 135,491 meals, the team identified thirty-four interpretable meal archetypes. They then conditioned a generative model and a portion predictor on them to build meals meeting US dietary targets. Generated meals landed much closer to recommended intake targets than the real meals behind them, while staying similar. Allowing one to three ingredient substitutions produced meals about ten percent more nutritious and, on average, nineteen to thirty-two percent cheaper. The authors see uses in programs and consumer apps, with clinical decision support a possibility after validation.
Catalog record
| CREATOR | Trevor Chan, Ilias Tagkopoulos |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |