Liver disease, a life-threatening malice, has become one of the most common diseases in recent years. Our goal is to identify the associated risks early enough through existing preconditions and make efficient predictions of liver disease using cutting edge machine learning models. The dataset is collected from UCI repository. Hybrid classifiers, developed by combining traditional classifiers with Bagging and Boosting methods like Gradient Boosting Boosting Method (GBBM), Random Forest Bagging Method (RFBM), Bagging Method (KNNBM), K-Nearest Neighbors, AdaBoost Boosting Method (ABBM), and Gradient Boosting Boosting Method (GBBM) have been used. K-nearest Neighbors performed the best with Testing Accuracy (TA) of 99.9%.