Farihah Farihah, Rofik Rofik

Detection of Breast Cancer Patient Mortality Status Using Machine Learning with SMOTE-Based Class Imbalance Treatment

Introduction

Detection of breast cancer patient mortality status using machine learning with smote-based class imbalance treatment. Detect breast cancer mortality risk using machine learning. This study employs SMOTE & Random Forest with GridSearchCV for early detection, achieving optimal prediction performance.

Abstract

Breast cancer is a disease with a high mortality rate, making early detection of patients at risk of death crucial for supporting medical decision-making. This study aims to develop a model for detecting patients at risk of death by comparing several machine learning algorithms. The research process included data collection, exploratory data analysis, data preprocessing, feature engineering, oversampling, modeling, and model evaluation. Data balancing was performed using SMOTE, and model optimization was conducted through hyperparameter tuning using GridSearchCV. The results show that Random Forest combined with GridSearchCV delivers the best performance in prediction, with an accuracy of 0.7491, precision of 0.2953, recall of 0.4634, an F1-score of 0.3608, and an ROC-AUC of 0.7030. This study demonstrates that the combination of Random Forest, SMOTE, and GridSearchCV is capable of optimizing the performance of the prediction model.

Full Text

Comments

You need to be logged in to post a comment.