DocumentCode :
3736521
Title :
Artificial neural networks: A solution for increasing the accuracy of regional traceability assessments
Author :
Mirela Praisler;Simona Constantin Ghinita;Atanasia Stoica Mandru
Author_Institution :
Department of Chemistry, Physics and Environment, "Dunarea de Jos" University of Galati, Galati, Romania
fYear :
2015
Firstpage :
1
Lastpage :
4
Abstract :
We are presenting a feasibility study regarding the use of Artificial Neural Networks for performing more detailed (regional) traceability assessments in the case of horticultural products. The challenge is related to the significant data variability and the need of fast data analysis and processing, especially in the case of fast perishable products. A case study performed for lovage (Levisticum Officinale) indicates that ANN may provide efficient and cost-effective automated regional traceability evaluations. This method yields a remarkable correct classification rate even for a simple (three layer) architecture and a training database built with a low number of physico-chemical properties.
Keywords :
"Artificial neural networks","Moisture","Feeds","Food products","Proteins","Training","Safety"
Publisher :
ieee
Conference_Titel :
E-Health and Bioengineering Conference (EHB), 2015
Print_ISBN :
978-1-4673-7544-3
Type :
conf
DOI :
10.1109/EHB.2015.7391556
Filename :
7391556
Link To Document :
بازگشت