The present results demonstrate that the correlation coefficient
of sugar-acid content, represented by complex components, was
higher than that of individual component correlation coefficients
in previous research. These findings indicate that the sugar and
acid content of Citrus could be estimated before harvesting, and
more accurate predictive component modeling is expected for
future PLS. For example, sample standardization could further
improve PLS modeling accuracy, as could increasing the number
of samples used in models, and the utilization of samples with
large differences in sugar-acid contents. PLS modeling is easy to
apply, costs relatively little, and enables the rapid prediction of
sugar-acid components when using the modeling described here