PERBANDINGAN KLASIFIKASI KANKER PAYUDARA MELALUI PENDEKATAN K-NEAREST NEIGHBORS DAN ARTIFICIAL NEURAL NETWORK

BENNY NURDIANTO, BENNY and JOKO TRILOKA, JOKO (2024) PERBANDINGAN KLASIFIKASI KANKER PAYUDARA MELALUI PENDEKATAN K-NEAREST NEIGHBORS DAN ARTIFICIAL NEURAL NETWORK. Other thesis, IIB Darmajaya.

[img] Text
COVER.pdf
Download (194kB)
[img] Text
ABSTRAK.pdf
Download (378kB)
[img] Text
HALAMAN PERSETUJUAN TESIS.pdf
Download (438kB)
[img] Text
HAL PENGESAHAN.pdf
Download (401kB)
[img] Text
DAFTAR ISI.pdf
Download (376kB)
[img] Text
BAB 1.pdf
Download (877kB)
[img] Text
BAB 2.pdf
Download (2MB)
[img] Text
BAB 3.pdf
Download (1MB)
[img] Text
BAB 4.pdf
Download (3MB)
[img] Text
BAB 5.pdf
Download (294kB)
[img] Text
DAFTAR PUSTAKA.pdf
Download (468kB)

Abstract

This study evaluates and compares the performance of K-Nearest Neighbors (KNN) and Artificial Neural Networks (ANN) algorithms in classifying breast cancer based on the Breast Cancer dataset, focusing on the use of feature selection using the Recursive Feature Elimination (RFE) method. The main objective of this research is to determine the accuracy, precision, recall, and F1 Score of each algorithm, both with and without feature selection, and identify the advantages and disadvantages of each approach. The results showed that the use of RFE significantly improved the performance of both algorithms. ANN with RFE achieved the highest accuracy of 98.64%, with precision of 99.31%, recall of 97.03%, and F1 Score of 98.13%, showing the superiority of ANN in breast cancer classification. In contrast, KNN with RFE also showed improved performance with 97.56% accuracy, 97.32% precision, 96.29% recall, and 96.71% F1 Score. Without feature selection, the performance of both algorithms decreased significantly, with ANN achieving 92.24% accuracy and KNN only 89.03%. These results confirm the importance of feature selection in improving model accuracy and stability, and show that ANN is superior in handling complex data compared to KNN. Based on these findings, ANN with RFE feature selection is recommended as a more effective model for breast cancer classification. Keywords: K-Nearest Neighbors (KNN), Artificial Neural Networks (ANN), Breast Cancer Classification, Breast Cancer Dataset, Feature Selection, Recursive Feature Elimination (RFE), Accuracy, Precision, Recall, F1 Score.

Item Type: Thesis (Other)
Subjects: Ilmu Komputer
eTheses
Divisions: Pasca Sarjana > Magister Teknik Informatika
Depositing User: Benny Nurdianto Nurdianto
Date Deposited: 31 Oct 2024 02:10
Last Modified: 31 Oct 2024 02:10
URI: http://repo.darmajaya.ac.id/id/eprint/18759

Actions (login required)

View Item View Item