A Machine Learningn based healthcare assistant that predicts the most likely disease based on symptoms provided by the user. The chatbot provides disease predictions and precautionary advice through an interactive web interface built with Streamlit.
- Predicts disease from symptoms using a trained Random Forest Classifier
- Displays precautionary steps associated with predicted disease
- Easy-to-use web interface built with Streamlit
- Reads and processes real-world healthcare datasets (symptom-disease and disease-precaution)
- Designed for local use without paid APIs or internet dependencies
| Layer | Tools/Frameworks |
|---|---|
| Programming | Python |
| ML Model | scikit-learn (RandomForestClassifier) |
| UI | Streamlit |
| Data Handling | pandas, NumPy |
| Visualization | Streamlit UI components |
https://www.kaggle.com/datasets/itachi9604/disease-symptom-description-dataset?resource=download
- Model Used: RandomForestClassifier from
scikit-learn - Input: One-hot encoded list of symptoms
- Output: Predicted disease (classification)
- Evaluation: Accuracy score, tested on holdout set
- Why RF?: Ensemble-based, handles imbalanced features well, robust to overfitting