Liver Cancer Classification Using Random Forest and Extreme Gradient Boosting (XGBoost) with Genetic Algorithm as Feature Selection
This paper will be focusing on the implementation of Genetic Algorithm as feature selection when applied to the widely used machine learning algorithm Random Forest and Extreme Gradient Boosting for cancer classification and the result will show that with 20% testing data, XGBoost with Genetic Al algorithm asfeature selection gives the highest accuracy score.
Abstract
Cancer has become a common cause of death in the world with approximately ten million deaths every year. Liver cancer is one of the common cancers that can happen to both men and women. Therefore, early detection is very important to help patients treating the cancer before it spreads to another organ. Machine Learning has been a huge help in many fields including the medical fields for classifying cancer data so the diagnosis can be more accurate. For that, feature selection is needed to increase the accuracy even more. This paper will be focusing on the implementation of Genetic Algorithm as feature selection when applied to the widely used machine learning algorithm Random Forest and Extreme Gradient Boosting (XGBoost) for cancer classification. The result will show that with 20% testing data, XGBoost with Genetic Algorithm as feature selection gives the highest accuracy score at 82%.