Domain-informed weight forecasting: leveraging behavioral and physiological sequences from wearables
Luping Cheng, Lu Wang, Gaolei Wang, Bo Lu, Bei Wu, Yang Xiao, Qian Huang · 2026
Read Domain-informed weight forecasting: leveraging behavioral and physiological sequences from wearables by Luping Cheng, Lu Wang, Gaolei Wang, Bo Lu, Bei Wu, Yang Xiao, Qian Huang (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-13269408.txt.
Description
Luping Cheng and co-authors build in 2026 a method to forecast a person's body weight a week ahead using only wearable and lifestyle data, never the scale itself. Frequent weigh-ins are impractical for many people, so the model reads fourteen days of behavior, physiology, and habits to project the next seven days of weight. Development used a synthetic long-term dataset, with people split into training and test groups before any data windows were cut.
A sequence-learning model beat tree-based rivals on the synthetic data and kept its edge when features were removed one by one. A small real-world test on sixteen people's five-month wearable logs showed the pipeline also runs outside synthetic data. The authors caution that week-scale shifts mostly reflect fluid and daily variation, not true fat change.
Catalog record
| CREATOR | Luping Cheng, Lu Wang, Gaolei Wang, Bo Lu, Bei Wu, Yang Xiao, Qian Huang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |