PREDICTION OF THE RIPENING TIMES OF MANCHEGO CHEESE USING MULTIVARIATE STATISTICAL ANALYZE - A PRELIMINARY-STUDY

Multiple linear regression, principal component regression, and partia l least squares (PLS) regression were used to calculate ripening time in standard Manchego cheeses based on water activity (Aw), pH, and oth er proteolysis parameters. The root mean square error of prediction by cross-validation was used to study the prediction accuracy of the mod els. PLS regression yielded the best predictions of ripening time and was used to predict the ripening times of eight commercial Manchego ch eeses.