Made by Masa Aladwan and Mohammad Moataz
Diabetes Platform is a web application for diabetes prediction and analysis. This project aims to predict the risk of diabetes using machine learning techniques in Python and analysis using Power BI. The model development process involved extensive research and work to create an accurate and reliable prediction system.
- Diabetes Risk Prediction: Utilizes a CatBoost machine learning model for predicting diabetes risk.
- FastAPI Server: Integrates a FastAPI server for efficient handling of API requests and responses.
- Data Storage: Stores predicted information in a PostgreSQL database.
- Power BI Dashboard: Provides insightful visualizations and analysis for diabetes data.
The pipeline for the Diabetes Platform is as follows:
- User Input: User inputs information through the web application.
- API Server: FastAPI server receives the information and inputs it into the ETL Python process.
- ETL Process: Python process performs data preprocessing, stores the information, and predicts the risk of diabetes.
- Storage: Predictions and information are stored in a PostgreSQL server.
- User Interface: User views the prediction results through the web application.
- Power BI Dashboard: Information is visualized and analyzed in Power BI dashboard.
- Web Development: HTML, CSS, and JavaScript for web application.
- API: FastAPI for handling API requests and responses.
- Machine Learning: CatBoost for diabetes risk prediction.
- Database: PostgreSQL for storing predicted information.
- Visualization: Power BI for creating insightful dashboards and reports.
- Association Rule: Power BI for creating insightful dashboards and reports.
-
Clone the repository:
git clone https://github.com/MohammadMoataz2/DiabetesPlatform.git
-
Install dependencies:
pip install -r requirements.txt
-
Run the application:
- index.html
- fast_api_server.py
-
Input your information to predict diabetes risk.
Contributions are welcome! Please fork the repository and submit a pull request with your changes.