Machine learning recommendation project
Medicine Recommendation System
This project uses a drug and symptom dataset to recommend possible medicines based on user input. It focuses on data preprocessing, matching symptoms, and generating relevant recommendations.
What I built
This project uses a drug and symptom dataset to recommend possible medicines based on user input. It focuses on data preprocessing, matching symptoms, and generating relevant recommendations.
Why I built it
Users need a simple way to explore possible medicine matches from symptom data, while keeping the system clearly positioned as an educational prototype.
My role
Prepared the dataset, implemented symptom matching logic, and created a testable interface for recommendation output.
How it works
- Input layer captures symptoms from the user.
- Preprocessing normalizes symptoms and maps them against dataset fields.
- Recommendation logic ranks possible medicine matches.
- Interface layer presents recommendations with clear prototype framing.
System architecture
- Input layer captures symptoms from the user.
- Preprocessing normalizes symptoms and maps them against dataset fields.
- Recommendation logic ranks possible medicine matches.
- Interface layer presents recommendations with clear prototype framing.
Screenshots and demo
Technical challenges
- Cleaning symptom labels so user input maps to dataset values.
- Avoiding overclaiming in a healthcare-related prototype.
- Keeping the interface simple enough for quick evaluation.
Results
- Symptom matching workflow
- Dataset-backed recommendation prototype
What I learned
- Recommendation systems need careful framing when domain risk is high.
- Transparent preprocessing improves trust in prototype outputs.
Explore the code
Explore the code, setup instructions, and technical documentation on GitHub.