Hyper-Parameter Tuning on Liver Disorder Prediction Models

DOI

https://www.doi.org/10.59256/ijrtmr.20260603020

Published In

Volume 6, Issue 3 (2026)

Published Online

May-June 2026

Pages

177-181

Article Type

Original Article

Check for updates

Public health is crucial study for any of the state & country to make precautionary measures to caution the public. In many situations, patients may require the blood to transfuse in case of surgery, trauma, chronic illness, blood disorders, etc. in such cases it is necessary to examine and ensure that the donor’s blood, whether donor’s liver is functioning properly, free from transmissible infections such as Hepatitis-C, Fibrosis, Cirrhosis and health status etc.., This paper presents the prediction of liver issues in blood donors based on the sample data set collected from medical records of 615 patients on thirteen attributes using various machine and deep learning techniques. To enhance the predictive performance of these models, applied Random search, and Bayesian optimization hyper-parameter optimization techniques and showed significantly improvement in accuracy for identifying individuals with various liver health diseases.