Data fusion using Multiple Factor Analysis coupled with non-linear pattern recognition (fuzzy k-means): application to Chenin blanc

Patterns in data obtained from wine chemical and sensory evaluations are difficult to decipher using classical statistics. Coupling fusion with machine learning techniques could assist solving these issues lead new hypotheses. The current study investigated the applicability of pattern recognition approaches for oenological applications. A sample set 23 Chenin blanc wines made young (< 35 years) old (> vines were analysed (recently bottled (Year 1) after two years storage 2)). Sensory (sorting) (NMR: nuclear magnetic resonance HRMS: high-resolution mass spectrometry) collected. Multiple factor analysis (MFA) was used fusion. Cluster performed by agglomerative hierarchical clustering (AHC) fuzzy k-means. Optimal cluster conditions found both methods cophenetic coefficient assess confidence fit. Given large number variables, models complex. Inconsistent patterns observed when varying conditions, indicating high similarity between samples. Overall, k-means resolved better than AHC and, coupled fusion, improved interpretation complex data.