Prediksi Prognosis Kanker Payudara Menggunakan Hybrid ANN-KNN

Abstract

Breast cancer is a disease with a high incidence and mortality rate worldwide. Determining patient prognosis is often challenging due to the complex clinical factors involved. This study aims to develop a model for predicting breast cancer prognosis using a hybrid Artificial Neural Network (ANN) and K-Nearest Neighbors (KNN) method. The dataset used is the Breast Cancer Wisconsin Prognostic (WPBC), obtained from the UCI Machine Learning Repository, comprising 198 patient data with 35 attributes. The research steps include data cleaning, label transformation, normalization using Min-MAX Scaler, feature extraction using ANN, and classification using KNN. The optimal K-value was determined by testing K=1, K=3, K=5, and K=7 using stratified K-fold cross-validation. Test results indicate that the highest accuracy of 84.62% was achieved in the fifth fold with K=7.