MedCare (Doceat) – AI Disease Prediction System
|Machine Learning Engineer (Personal Project)
Nigeria, Nigeria, Nigeria
Summary
Designed and deployed an AI-powered disease prediction system to assist healthcare staff in assessing likely illnesses from early symptoms, improving hospital triage workflows.
Highlights
Generated a synthetic healthcare dataset of over 5,000 records and engineered features (age, gender, fever, cough, headache, body pain) using Label Encoding for multi-class prediction.
Trained and evaluated a Decision Tree Classifier, achieving high classification accuracy after dataset balancing and model tuning.
Successfully predicted multiple diseases, including Malaria, Flu, Typhoid, and Common Cold, demonstrating robust model performance.
Built a clean, multi-step Streamlit web application with improved UI/UX, enabling real-time disease prediction for healthcare screening use cases.
Delivered a scalable AI healthcare assistant, fully deployed on Streamlit Cloud, optimized for hospital triage workflows.