Health data is used in a wide range of applications to assess and analyze the potential risks linked to an individual's health status. Various analytical techniques are used to evaluate risk factors, enabling more accurate predictions and informed decision-making. Such applications are essential for the proactive management of health, facilitating the identification of potential health issues before they become critical and improving overall healthcare outcomes. Fuzzy-rough set-based feature selection improves the accuracy and efficiency of data analysis, particularly when applied to large datasets. This paper presents a machine learning approach for heart risk analysis, leveraging feature selection techniques applied to real-valued attributes within a sample dataset using fuzzy-rough set-based redact theory. The proposed approach focuses on identifying and selecting the most relevant features from the dataset that are crucial for making informed decisions and accurately predicting an individual’s health risk factors. By employing this methodology, we aim to enhance the effectiveness and precision of health risk assessments through the reduction of unnecessary or redundant data, thereby improving overall analytical outcomes.