Discriminant analysis of countries growing wakame seaweeds: a preliminary comparison of visible-near infrared spectra using soft independent modelling, Randomforests and classification and regression trees

摘要

Concerns about the labelling of fisheries products at production centres is now increasing. After the outbreak of bovine spongiform encephalopathy (BSE) occurred, problems relating to clam or wakame and other products have attracted more attention than before. We developed a scientific method for discriminating the production countries of wakame seaweeds using soft independent modelling for class analogy (SIMCA). However, the validation of the obtained SIMCA model was not very good. Therefore, we tried one of the non-parametric classification methods, Randomforests, based on classification and regression trees (CART). We found that Randomforests could be a potent classification method for visible and near infrared spectra. We suggest a preliminary interpretation for the basis of classification conditions by using CART.