Liver Disease Identification Based on Machine Learning Algorithms

Liver disease became a worldwide issue that is more required to be solve. Early detection of disease is extremely important for deriving the best treatment during the patient's treatment process and saving lives. Using several Machine Learning (ML) approaches, this research attempts to predict and diagnose liver illness. We utilised a Kaggle standard dataset that included records of patients with and without liver disease. These records were collected in Andhra Pradesh's North East area. The collection contains 441 male patient records and 142 female medical records. Furthermore, the dataset comprises ten characteristics and one target that differentiate between liver disease (1) and non-liver disease (0). The evaluation results depend upon the accuracy values indicated that the results achieved are 76.07%, 74.36%, 71.79%, 70.09%, and 80.34% for LR, SVM, GB, DT, and RF, respectively. The results indicated that after applying Principal Component Analysis (PCA) as a feature selection algorithm the RF achieved higher accuracy 80.37% than other recent approaches.