Authentication and Provenance of Walnut Combining Fourier Transform Mid-Infrared Spectroscopy with Machine Learning Algorithms.

ABSTRACT: Different varieties and geographical origins of walnut usually lead to different nutritional values, contributing to a big difference in the final price. The conventional analytical techniques have some unavoidable limitations, e.g., chemical analysis is usually time-expensive and labor-intensive. Therefore, this work aims to apply Fourier transform mid-infrared spectroscopy coupled with machine learning algorithms for the rapid and accurate classification of walnut species that originated from ten varieties produced from four provinces. Three types of models were developed by using five machine learning classifiers to (1) differentiate four geographical origins; (2) identify varieties produced from the same origin; and (3) classify all 10 varieties from four origins. Prior to modeling, th

SUBMITTER: Zhu H 

PROVIDER: S-EPMC7662659 | biostudies-literature | 2020 Oct

REPOSITORIES: biostudies-literature