FEBBI YUSTITIA AKSARI, 101614153042 (2018) PERBANDINGAN KETEPATAN KLASIFIKASI KASUS KANKER SERVIKS MENGGUNAKAN RANDOM FOREST DAN SUPPORT VECTOR MACHINE (SVM) (Studi di Yayasan Kanker Wisnuwardhana Surabaya). Thesis thesis, Fakultas Kesehatan Masyarakat.

Preview	Text (abstrak) abstrak.pdf Download (21kB) | Preview
	Text (full text) full text.pdf Restricted to Registered users only until 11 October 2021. Download (1MB)

Abstract

Random Forest (RF) and Support Vector Machine (SVM) are computational-based statistical method. The former is used for classification cases, while the former is used for prediction and classification cases. SVM is a statistical method that can also identify cases of cervical cancer based on the results of examination by considering the factors of identification (risk factors). Meanwhile, RF is one of the methods used for classification by building many classification trees. The general objective of this study was to compare RF and SVM performances in determining the classification of risk factors (age of first marriage, number of children, type of contraceptves, duration of contraceptive use, and patient age) affecting the incidence of cervical cancer. The type of this research was unobstrusive research. The sample was the total population that had inclusion criteria, i.e. female patient who had performed papsmear at Wisnuwardhana Cancer Foundation Surabaya in January-December 2017. SVM analysis results show that 1-APER was 0.995 or in other words, the right sample data were classified 99.5%. Sensitivity of SVM was 0.997 and the specificity was 0.988. Furthermore, the RF analysis results show that the calculation results of 1-APER was 1.00 or in other words, the right sample data were classified 100%. The sensitivity was 1.00 and the specificity was also 1.00. The conclusion of this research is that all measures of accuracy that include 1-APER, sensitivity, specificity, and G-means in RF method is better than the support vector machine method. This indicates that RF method is better at classifying the cases of cervical cancer at Wisnuwardhana Cancer Foundation Surabaya.

Actions (login required)

View Item