Description
Flavor pairing using machine learning algorithms is a modernist technique that leverages computational analysis of volatile compound profiles to predict compatible food combinations.
Technical
The method relies on gas chromatography–mass spectrometry (GC‑MS) to profile volatile compounds such as terpenes, phenols, sulfur‑containing molecules, esters, and lactones. Machine‑learning models compute similarity scores (e.g., cosine similarity) between ingredient vectors; scores above a threshold (≈0.7) suggest compatible pairings. Recommended cooking at 80–90 °C for 5–10 min volatilizes aromatic esters and lactones, enhancing perceived similarity.
Science
Primary Reaction
Analysis of volatile compound profiles and identification of shared aromatic compounds
Sensory Profile
Aroma ()
Origin & History
Civilization