Machine learning-based disease prediction system using symptom analysis

Early and accurate diagnosis is critical for effective disease prevention and treatment. Traditional diagnostic methods may be insufficient for serious medical conditions. This research presents a machine learning (ML) based disease prediction system that can assist in accurate diagnosis. We developed a comprehensive disease prediction framework using multiple ML algorithms to analyze a dataset containing over 230 diseases. The system predicts potential diseases based on an individual’s symptoms, age, and gender. Among the evaluated algorithms, the weighted K-Nearest Neighbors (KNN) model achieved the highest accuracy at 93.5%. Our diagnostic model can serve as an automated preliminary screening tool for early disease identification, enabling timely treatment initiation and potentially saving lives, particularly in regions with limited healthcare resources or during public health emergencies.