Predicting Hepatitis C Virus Using Machine Learning

Abstract

A virus that causes called hepatitis C damages and incites the liver. Enlargement triggered by inflammation happens whenever bodily parts are hurt and diseased. Body parts could be harmed by irritation. Most of the world's cases of hepatitis C are found in countries like Egypt. According to estimates, there are 3–4 million new cases each year, making it a public health concern that needs to be addressed with treatment and screening programs. In recent years, diagnostics-based technology has dramatically advanced. AI systems are able to create diagnostic models from a wide range of unusually complicated structures. I employed machine learning methods in this research to assess the presence or absence of the hepatitis C virus. The availability of very accurate risk prediction models would make it easier to identify those who need more intensive monitoring and treatment ahead of time. Individuals with chronic hepatitis C (CHC), the most common cause of cirrhosis worldwide, would benefit most from risk prediction models. Despite the availability of efficient antiviral treatment for CHC, the disease has yet to be eradicated. There are six ML algorithms are applied that are used Logistic Regression, K Neighbors Classifier, Random Forest Classifier, Cat Boost Classifier, Decision Tree Classifier, and Gradient Boosting Classifier. The gradient-boosting algorithm generated a number of the top 6 results. The answer is 94.31 percent. I also find out the confusion matrix of these algorithms and use a correlation matrix to calculate the following numbers: Individuals who are suspicious 5402. 75 healthy patients, 61.30% of whom are men and 38.70% of whom are women.