Different supervised and unsupervised classification approaches based on visible/near infrared spectral analysis for discrimination of microbial contaminated lettuce samples: Case study on E. coli ATCC

•Safety assessment of lettuce based on E. coli using the Vis/NIR spectra is possible.•SIMCA & SVM can be more reliable than PCA & HCA to distinguish unsafe samples.•Microbial population amount cause a complex process in lettuce.•Chemical component of lettuces with E. coli concentration have a non-linear relationship.•PLSDA provides a robust qualitative model accurately classified lettuce samples.