Identification of Tuna and Mackerel Based on DNA Barcodes using Support Vector Machine
Abstract: Tuna and mackerel
are important fish in Indonesia that have great demand in the community and contain
good nutrients for health. Many of the processed products have been faked
including processed fish, by replacing the content of products that have high
sales value to other lower price one. For ensuring food safety, fraudulent
should be prevented by identifying the content of refined product. In this
research, we implemented support vector machine (SVM), one of the popular
methods in machine learning, to yield a model for identifying the content of
refined product based on DNA barcode sequences. The feature extraction of DNA
barcode Sequences was conducted by calculating k-mers frequency of each
sequences. In this study, we used trinucleotide (3-mers) and tetranucleotide
(4-mers). These features were inputted to SVM to classify and identify whether
the DNA barcode sequences belong to the class of tuna, mackerel, or other fish.
The evaluation results showed model SVM was able to perform identification with
the accuracy 88%.
Author: Mulyati, Wisnu Ananta
Kusuma, Mala Nurilmala
Journal Code: jptkomputergg160192
