Patient Risk Identification Using Machine Learning

Abstract: These days, hospitals and clinics keep a lot of patient information on computers. Trying to manage all these records by hand gets really hard as more and more patients come in every day. Doctors often have to look at medical reports, lab results, and a patient's past quickly. Because of this, healthcare places are slowly moving towards smart systems that can help them analyze things and make predictions faster.
This project is about finding out patient health risks using computer programs called machine learning algorithms. The system figures out if a patient is at low or high risk by looking at health details like blood pressure, sugar level, cholesterol, heart rate, and age. While we were building it, we tried and tested different algorithms, including Logistic Regression, Decision Tree, Support Vector Machine, and Random Forest.

We built the application using Python and gave it a simple screen so people could easily use it. After trying out the different models, the Random Forest algorithm gave us better results than the other methods we used. This system can help medical staff with their first look at a patient's case and potentially help them spot health risks sooner.

Keywords: Patient Risk Prediction, Machine Learning, Healthcare Analysis, Disease Prediction, Random Forest