Cluster analysis and artificial neural networks multivariate classification of onion varieties

J Agric Food Chem. 2010 Nov 10;58(21):11435-40. doi: 10.1021/jf102014j. Epub 2010 Oct 15.

Abstract

Eight cultivars of different colored onions (white, golden, and red) were evaluated for fresh bulbs cultivated and grown under the same environmental and agronomical conditions. Cluster analysis and principal component analysis, based on different flavonoids, total phenols, and pungency, data showed that the onions were not clustered according to variety (genetic similarity degree), whereas the color was the variable with the highest influence, ranging between 50 and 70%. Artificial neural networks were applied to study the possibility of discriminating among onion varieties. Characterization of the onion according to variety and procedence of the seeds was around 95-100%. Samples belonging to the Carrizal Alto procedence had an incorrect classification for 25% of the data.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Cluster Analysis
  • Discriminant Analysis
  • Neural Networks, Computer
  • Onions / chemistry*
  • Onions / classification*
  • Phenols / analysis
  • Plant Extracts / analysis
  • Plant Roots / chemistry

Substances

  • Phenols
  • Plant Extracts