Potato quality assessment by monitoring the acrylamide precursors using reflection spectroscopy and machine learning

Acrylamide formation is nowadays one of the major concerns potato-processing agriculture industry. We investigate use broadband reflection spectroscopy (400–1700 nm), in combination with machine learning, to optically classify raw potatoes inducing different levels acrylamide after frying, covering concentrations between 200 ppb and 2000 ppb. Using full spectral range, we obtain a correct classification dataset samples using 10-fold cross-validation, while applying Linear Discriminant Analysis Extreme Learning Machine. To reduce amount data increase processing speeds, sequential feature selection search was performed identify critical wavelengths (450 nm, 488 504 783 808 1310 1319 nm 1342 nm) that enable performances exceeding 92% when Analysis. therefore demonstrate non-destructive identification unsuited for French fries production, enabling food safety, limiting waste.