Symptom based disease prediction using machine learning

This project presents a system for predicting diseases based on given symptoms using machine learning techniques. The dataset used in this work is synthetic and consists of multiple symptom-based features represented in numerical form. The system applies various classification algorithms such as Decision Tree, Random Forest, XGBoost, Voting Classifier, and Stacking Classifier to identify patterns between input features and target classes. The goal is to improve prediction accuracy by combining multiple models and selecting the most effective approach.

The system is designed with a structured flow that includes user interaction modules such as registration, login, and classification. Once the user inputs the symptoms, the trained model processes the data and predicts the corresponding disease category. After prediction, the system provides a treatment approach using acupressure techniques through an animated representation, along with precautionary measures and recommendations.

The implementation uses a simple web interface supported by a backend framework, ensuring smooth data handling and model integration. The project demonstrates how machine learning models can be effectively used for classification tasks and enhanced with additional supportive features for better understanding and usability. The results highlight the importance of ensemble methods in improving prediction performance.

Keywords: Machine Learning, Classification, Symptoms, Disease Prediction, Ensemble Methods, Random Forest, XGBoost, Voting Classifier, Stacking, Acupressure