Comparing support vector machine and random forest to predict ovarian cancer

Annisa Wardhani, Zuherman Rustam, Sri Hartini, Glori Stephani Saragih, Shafira Habibah

Abstract


The third most common female cases of cancer in Indonesia is Ovary Cancer. Ovary cancer or ovarian cancer, affecting one or both ovaries, is a heterogeneous disease. The factors which increase risk of ovarian cancer are more pregnancies, early menstruation and late menopause, obesity, family history, and BRCA1 or BRCA2 mutation. Ovarian cancer can be diagnosed by physical test, pelvic test, transvaginal ultrasound test, x-ray, MRI scan, CT scan, or CA 125.  Patients of ovarian cancer who are treated should do regular checkup, because it may come back. This research uses data of ovarian cancer examination result from RS Al Islam Bandung hospital. The method that used are Support Vector Machine (SVM) and Random Forest method. SVM is the classification method that uses discriminant hyperplane, dividing to classes. Then, Random Forest is a tree-based ensemble depending on a collection of random variables. SVM and Random Forest are applied to predict ovarian cancer data, then the results of each methods are compared. Random Forest method is better than SVM with the accuracy value is 100%. 

Keywords


Classification; Confusion matrix; Ovarian cancer; Random forest; Support vector machine (SVM)



DOI: http://doi.org/10.11591/ijai.v10.i2.pp%25p

Refbacks

  • There are currently no refbacks.


View IJAI Stats

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.