KLASIFIKASI PENYAKIT KANKER SERVIKS MENGGUNAKAN ALGORITMA FORWARD SELECTION DAN K-NEAREST NEIGHBOR (KNN)

Dedi Arbain, Dedi Arbain and Sriyanto, S.Kom., MM., Ph.D, Sriyanto (2025) KLASIFIKASI PENYAKIT KANKER SERVIKS MENGGUNAKAN ALGORITMA FORWARD SELECTION DAN K-NEAREST NEIGHBOR (KNN). Other thesis, IIB Darmajaya.

[img] Text
COVER.pdf - Other
Download (137kB)
[img] Text
ABSTRAK.pdf - Other
Download (115kB)
[img] Text
PENGESAHAN.pdf - Other
Download (573kB)
[img] Text
DAFTAR ISI .pdf
Download (150kB)
[img] Text
BAB 1.pdf
Download (316kB)
[img] Text
BAB 2.pdf
Download (422kB)
[img] Text
BAB 3.pdf
Download (313kB)
[img] Text
BAB 4.pdf
Download (258kB)
[img] Text
BAB 5.pdf
Download (118kB)
[img] Text
DAFTAR PUSTAKA.pdf
Download (183kB)
[img] Text
LAMPIRAN.pdf
Download (2MB)

Abstract

The collection of past medical records of cervical cancer patients in health facilities was not accompanied by a process of extracting knowledge and information. The use of data techniques mining could be implemented into a system that can predict cervical cancer. The research focused on collecting data on medical diagnoses of patients undergoing pap smears. The dataset consisted of 36 variable parameters of information, habits, and medical records of 858 patients. Some patients may refuse to answer some questions because of privacy reasons (missing values). This dataset contains four Boolean-type target variables: Biopsy, Cytology, Hinselmann, and Schiller. Each target is data about the patient's test results, such as detecting cancer cells in the patient's cervix. The algorithm used for Cervical Cancer Classification is Forward Selection and knearest neighbors (K-NN). The testing was carried out on the Forward Selection and k-nearest neighbors (K-NN) algorithms by using the Confusion matrix formula. Test results for the algorithm were used as much to show that the Forward Selection and K-nearest algorithm neighbors (K-NN) have the highest accuracy of 97.35% using Hinselmann variables. This showed that the performance achieved by each algorithm used allowed the cervical cancer prediction system to be used to support clinical decisions in new patients. Keywords: Cervical Cancer, Forward Selection, K-NN, Classification, Performance, Matrix Confusion

Item Type: Thesis (Other)
Subjects: Ilmu Komputer
eTheses
Divisions: Pasca Sarjana > Magister Teknik Informatika
Depositing User: DEDI DEDI ARBAIN
Date Deposited: 09 Jan 2025 07:38
Last Modified: 09 Jan 2025 07:38
URI: http://repo.darmajaya.ac.id/id/eprint/19149

Actions (login required)

View Item View Item