Klasifikasi Fragmen Metagenome Menggunakan KNN dan PNN dengan Ekstraksi Fitur Gray Level Co-occurrence Matrix (GLCM) pada Variasi Panjang Fragmen

View/Open
Date
2014Author
Dhira, Muhammad
Kustiyo, Aziz
Kusuma, Wisnu Ananta
Metadata
Show full item recordAbstract
Bioinformatics is a field of study which is developing rapidly in Indonesia. The main focus of this study is to classify metagenome fragment into some taxonomies using GLCM that has 13 features. The training data used in the classification process are 50 organisms with 5-fold cross validation. The DNA combination sequences inside a fragment can be seen as a 1xN image, where N is the fragment’s length in two-dimension’s matrix. The result from those features will be classified using PNN and KNN. The research shows that the accuracy percentage with all lengths variety, including 200 bp, 1 Kbp, 3 Kbp, and 10 Kbp, are 100%. It can be stated that the variety fragment’s length does not affect the accuracy. In addition, it can be concluded that GLCM feature extraction method can be prospectively implemented for classifying metagenome fragment.
Collections
- UT - Computer Science [2335]