KLASIFIKASI SEL TUNGGAL PAP SMEAR BERDASARKAN ANALISIS FITUR DAN ANALISIS TEKSTUR TERSELEKSI MENGGUNAKAN CORRELATION-BASED FEATURES SELECTION BERBASIS DECISION TREE J48

Asti Herliana, Dwiza Riana

Abstract


Abstrak

 This research presents the texture classification of single cells Pap Smear. The single cells of Pap Smear have many kind of texture parameter that have been discovered by giffary, et al on 2012 research.  By using the Correlation-based Features Selection (CFS) to select the texture parameter that produce correlation135, energy0, deviation and brightness as the best parameter to increase the classification result. In this research, the best parameter of texture was combined with Kerne_A and Cyto_A from the features parameter that has been discovered from Martin(2003) and Jantzen et al (2005). By using the method of Decision Tree Classifier to the six selected parameter (Correlation135, Energy0, Deviation, Brightness, Kerne_A and Cyto_A)result the accuracy about 90% for the two classes and about 67,87% for the seven classes.

 


Full Text:

Untitled

Refbacks

  • There are currently no refbacks.