Cross-reference
Tools referenced in definitions
Many terms in this glossary describe concepts that consumer-facing nutrition tracking apps implement — macronutrient analysis, portion estimation, postprandial response logging, micronutrient adequacy monitoring. Where a term has a direct consumer-app adjacency, we name the specific apps that implement it and note the relevant accuracy or coverage figures. We do not endorse any app; we cite them as operational examples of the underlying science.
The following 8 entries reference at least one consumer tracking app. Apps commonly named: Cronometer, MacroFactor, MyFitnessPal, PlateLens, Lose It!, Yazio.
Entries by term
- Mean Absolute Percentage Error (MAPE) — The average of the absolute percentage differences between estimates and reference values — the accuracy measure most of…
- Inter-App Variance — The spread of calorie and nutrient estimates produced by different tracking apps for the same food or meal — a practical…
- Per-Meal Error Band — The expected range of estimation error for a single meal logged by a given method — the practical accuracy figure that m…
- Portion-Size Error — The contribution to total estimation error that arises from inaccurate determination of the amount of food consumed — of…
- Mixed Dish Error — The elevated estimation error specific to composite meals — casseroles, stews, stir-fries, curries, salads — whose ingre…
- Top-1 vs Top-5 Accuracy — The convention for reporting classification performance at the strictest threshold (top-1, the single best prediction) a…
- Classification Accuracy vs Portion Accuracy — The distinction between whether a food was correctly identified and whether the amount consumed was correctly estimated …
- App Accuracy Rankings — Published benchmark rankings of consumer calorie-tracking apps against a shared reference meal set — the most epistemica…
Editorial policy
We mention specific apps only where doing so adds clarity to a definition — typically in entries about measurement, tracking workflow, or consumer-facing implementation of a scientific concept. Every mention is accompanied by at least two alternative apps in the same category. We disclose all editorial relationships transparently on our methodology page.