Nonlinear Canonical correspondence Analysis: Description of the data of Coffee

The formulation of coffee blends is paramount importance for the industry, as it provides product with an expressive ability to compete in market and adds sensory attributes that complement consumption experience. Through redundancy analysis canonical correspondence analysis, possible study relationships between a set notes different proportions variety through multivariate linear regression models. However, unrealistic assume such responses are given linearly relation blends, since some species have greater weight evaluation (quadratic terms) effect mixtures (term interaction). With this motivation, work aims propose use nonlinear polynomial evaluate acceptance varieties according scores by evaluators. Finally, concluded there were gains percentage total explained variance models classic