An advanced deep learning approach for predicting liver cirrhosis

Abstract

This study provides an advanced deep learning method for predicting liver cirrhosis using a Kaggle dataset of 615 entries with a target attribute "Liver Cirrhosis Status" that classifies answers as 'Yes' (75 cases) or 'No' (540 cases). Complete Data collection, Preprocessing, Model selection, Training, and Evaluation are all part of the suggested methodology. The trial findings reveal that the Artificial Neural Network (ANN) outperforms other categorization algorithms with its high accuracy of 98.01%. This discusses the model' extraordinary ability to identify detailed patterns in the medical dataset, proving its potential for accurate liver cirrhosis prediction. The ANN's achievement shows the utility of advanced deep learning techniques in medical testing, particularly for complex problems such as liver cirrhosis prediction. The group methods Random Forest Classifier and Ada Boosting, as well as SVC's unfair capabilities, all performed well, suggesting they are suitable for capturing subtle relationships within the dataset. These findings add to the growing body of knowledge about the use of complex neural networks created in healthcare, paving the way for better patient outcomes through early and exact prediction of liver cirrhosis. The findings of the study have important implications for continuing efforts to improve medical diagnostic capacities through the use of modern artificial intelligence technologies.