PENERAPAN DECISION TREE C4.5 SEBAGAI SELEKSI FITUR DAN SUPPORT VECTOR MACHINE (SVM) UNTUK DIAGNOSA KANKER PAYUDARA

Abstract

In the field of medicine, the use of data mining has a quite important and evolutionary role that can change the perspective of doctors, practitioners and health researchers in the process of detecting breast cancer in a patient. There are 2 classification applications in it, namely the process of diagnosing (diagnosing) cancer cells that distinguishes between tumors (benign cancer) or malignant cancer and prognosis (prognosis) to determine the possibility of reappearance of cancer cells in patients who have been operated on in the future. Data mining aims to describe new findings in the dataset and explain a process that uses statistical, mathematical, artificial intelligence, and machine learning techniques to extract and identify useful information and related knowledge from the database.

Classification with data mining can be done using several methods, namely Decision Tree, K-Nearest Neighbor, Naive Bayes, ID3, CART, Linear Discriminant Analysis, etc., which certainly have advantages and disadvantages of each. But in this study, the author focuses on the classification of data mining using the Support Vector Mechine and Deccision Tree algorithms.

This study will analyze the Breast Cancer Wisconsin Original data set obtained from the UCI Machine Learning Repository (repository of research data) to classify breast cancer malignancies. This time the author correlates between the Decision Tree classifier algorithm which has good ability to process large databases as a feature selection, then with a proper and relevant SVM Method used in analyzing and diagnosing breast breast cancer patients because it has accurate results for existing problems and several bases .

Keywords— Data Mining, diagnosis, Decision Tree, SVM Method