Jurnal Lebesgue (Apr 2024)
PENDEKATAN REGRESI LOGISTIK BINER DAN REGRESI LOGISTIK BERSTRUKTUR POHON DALAM ANALISIS DIAGNOSIS KANKER PAYUDARA
Abstract
The purpose of this research is to determine the factors that have a significant effect on the types of malignant and benign breast cancer using the binary logistic regression approach and classify the factors that have a significant effect on the types of malignant and benign breast cancer using the regression method. tree-structured logistics on breast cancer data from H. Adam Malik Hospital Medan 2021. This research is a type of applied research by collecting secondary data from the H Adam Malik General Hospital Medan. The research variables used are the response variable (Y) and the predictor variable (X). In the response variable there are also two categories, namely benign breast cancer patients and malignant breast cancer patients. While the predictor variables in this study were age, age at menarche, age at menopause, age at parity, use of hormonal contraception, genetics, number of children, breastfeeding period, and mammary fibroadenosis. This study uses the binary logistic regression method because it is a good method in determining factors. In addition, the tree-structured logistic regression method is also used because it is a good method in determining factor classification with a good level of accuracy for all data. These two methods can make it easier for researchers to calculate and determine the factors that have a significant effect on the type of malignant breast cancer and benign breast cancer using all the available variables. The results of this study obtained an accuracy value of 94% for the binary logistic regression method and 97.74% for the tree-structured logistic regression method. Because it has an accuracy value of almost 100%, this means that the binary logistic regression method and tree-structured logical regression applied to this research problem are good enough
Keywords
