This document presents a study on using machine learning algorithms to predict minimum health insurance premium amounts based on individuals' health parameters. The study analyzes various health-related attributes like age, diabetes status, surgeries, etc. from datasets to train regression models. Random forest regression is identified as the best performing algorithm with an accuracy of around 85-90%. The proposed system aims to help individuals choose more appropriate insurance amounts by considering a wider range of health factors compared to existing solutions. It describes the system architecture involving data collection, model training, and premium prediction when users input their health details.