Comparison between visual assessments and different variants of linear discriminant analysis to the classification of Raman patterns of inkjet printer inks

•Micro Raman spectroscopy is a suitable technique to detect in situ chemical information from cyan, magenta and yellow microscopic dots of inkjet printed documents.•With the aim to produce investigative leads, among the three tested variants of linear discriminant analysis (PCA + LDA, PLSDA, and sparse LDA), sparse LDA showed the best performances to the classification of the individual dot colors.•The three LDA variants performed equally when the data from all the three colored components were analyzed jointly.•Although the high classification accuracy values obtained with the tested LDA variants, visual comparisons of Raman profiles are still superior to detect differences based on minor peaks.