RANDOM FORESTOPTIMISINGLIVER DISEASEPREDICTION THROUGHDIFFERENT DATABALANCINGTECHNIQUES

Authors

  • Kapilavai Purushotham Raju Author

Keywords:

imbalance, skewness, outliers

Abstract

A prevalent condition, apart from heart attacks,
which take many lives, is liver disease. Due to the
fact that liver disease is usually detected late and
causes mortality. The number of liver patients is
increasing due to a variety of circumstances, such as
drinking contaminated water, inhaling toxic gas,
abusing alcohol excessively, and other behaviours
that might affect health parameters. By using
machine learning prediction models, these health
parameters may be utilised to predict liver ailments
in their early stages. The machine-learning model in
thisstudy is developed using the Indian Liver Patient
Dataset (ILPD), which is based on Indian patients.
Several preprocessing methods are used with the
Random Forest (RF) algorithm to forecast the
illness. The data collection is examined using
univariate and bivariate analysis to check for
imbalance, skewness, and outliers. After that,
appropriate algorithms are used to remove outliers,
and different techniques for oversampling and
undersampling are applied to balance the data. Grid
search and feature selection are used in conjunction
with hyper parameter adjustment to furtherrefine the
model. The final model has 100% accuracy and
performs verywell across all criteria

Downloads

Published

30-04-2024

How to Cite

RANDOM FORESTOPTIMISINGLIVER DISEASEPREDICTION THROUGHDIFFERENT DATABALANCINGTECHNIQUES. (2024). International Journal of Engineering Research and Science & Technology, 20(2), 236-242. https://ijerst.org/index.php/ijerst/article/view/278