Patent Number: 6,324,531

Title: System and method for identifying the geographic origin of a freshcommodity

Abstract: The detection method includes generating a plurality of neural networkmodels. Each model has as a training set a data set from a plurality ofsamples of a commodity of known origins. Each sample has been analyzed fora plurality of elemental concentrations. Each neural network model ispresented for classification a test data set from a plurality of samplesof a commodity of unknown origins. As with the training set, the sampleshave been analyzed for the same plurality of elemental concentrations.Next a bootstrap aggregating strategy is employed to combine the resultsof the classifications for each sample in the test data set made by eachneural network model. Finally, a determination is made from the bootstrapaggregating strategy as to a final classification of each sample in thetest data set. This final classification is indicative of the geographicalorigin of the commodity. The system includes software for generating theneural network models and a software routine for performing the bootstrapaggregating strategy.

Inventors: Anderson; Kim A. (Troy, ID), Smith; Brian (Troy, ID), Magnuson; Bernadene (Moscow, ID)


International Classification: G06N 3/04 (20060101); G06N 3/00 (20060101); G06K 9/62 (20060101); G06N 003/02 ()

Expiration Date: 11/27/2014