Performance Evaluation of Pap Smear Cell Image Classification Using Quantitative and Qualitative Features Based on Multiple Classifiers
Dwiza Riana, Aniati Murni · Repository Universitas Bina Sarana Informatika (RUBSI) · 2009
image class. Three classes of which are normal on Pap smear cell image classification using qualitative class categories that include: Normal Superfi features, as an effort to improve the results of the Normal Intermediate, and Normal Colu previous study using quantitative features especially in Abstract- This paper presents the results of a study whereas the other four classes arecategori the performance of the case using a/l(7)-c1ass category abnormal cells that include: Mild (Light) Oypl of diseases. Three good c1assil1ers have been chosen Moderate Dysplasia, Severe Dysplasia and Carcin and they include the Nai've Bayes (NB), Multi-Layer Pcrccptrons (MLPs), and Decision Tree learning In Situ (2). For the classification process, JantzCll algorithm (J48). Twenty quantitative features and sevenl al. (1) and Martin (2) have used least square met categories of Pap smear cell image class have been used a simple classifier. Norup (3) compared nearest in this study. Three classes of which are normal cell gravity center and neuro fuzzy inference meth image class categories that include: Normal Superficial, Pap smear cell classification. The study reported Normal Intermediate, and Normal Columnar, and the the nearest class gravity center performed better other four classes are categories of abnormal cell image class that include: Mild (Light) Dyplasia, Moderate neuro fuzzy inference. Dysplasia, Severe Dysplasia and Carcinoma In Situ. If Amalia el 01. (4) has used the quanti The conversion of quantitative to qualitative features features and Adaptive :<etwork-based Fuzzy Inre and the image classification process were done 'vith the System (At'\1FIS) classifier, while Giriel at. (:) aid of the Weka software. The experimental study used qualitative features and association rules, shows that in aJl(7)-c1ass classification based on the qualitative features and using the three classifiers ,vith classifY a Pap smear cell image. Their experim