A Comprehensive Evaluation of Ensemble Machine Learning Algorithms for Sepsis Prediction: A Comparative Study

Sepsis is still a big problem in health care, and it can be hard to spot and treat on time. Traditional ways of spotting sepsis aren't always good enough to make a quick diagnosis, so new ways are being looked into. Since it becomes a big health issue, hospital expenditures is increasing with time span. Machine learning, branch of Artificial Intelligence, has become a useful way to find sepsis. This study focuses on detecting sepsis using several machine learning models. There are six different models used, including logistic regression(LR), K-Nearest Neighbors (KNN), Random Forest (Random Forest), LGboost, Adaboost and XGBoost. The performance of the models is assessed using various evaluation metrics and comparison has been made. The experimental results demonstrate that the Random Forest, LGBoost and XGBoost model are the best imputation approach for predicting the disease's early stage with an accuracy of 77.5%,74.2% and 74.2% respectively. These are suitable for machine learning applications and is especially useful when dealing with complex and high-dimensional data-sets.