Raheesp/ML-XGBOOST-Predict-Diarrhea-Risk

The Effects of Community Health Worker Visits and Primary Care Subsidies on Health Behavior and Health Outcomes for Children in Urban Mali 2014 # 🧪 Diarrhea Risk Prediction Web App This project is a **Flask-based web application** that uses a trained machine learning model to predict the **risk of diarrhea in children** based on household and healthcare-related information. The data used is from a real-world health study conducted in **Urban Mali, 2014**. ## 📊 Project Overview The application allows users to input various child and household health details and get a **prediction of diarrhea risk (High/Low)** along with a **confidence score**. The model was trained using **XGBoost** and **RFE (Recursive Feature Elimination)** to select the most relevant features. --- ## 🚀 Features - 🧠 Machine learning model with feature selection (RFE) - 🎯 Risk prediction (High Risk / Low Risk) - 📈 Visual confidence score using a dynamic progress bar - 📄 Form inputs with descriptions for easy understanding - 💡 Animated transitions using CSS - 💻 Fully responsive Bootstrap UI --- ## 📁 Folder Structure ``` ├── app.py # Flask backend ├── templates/ │ └── index.html # Main frontend HTML file ├── static/ │ └── background/ │ └── image.jpg # Background image for the UI ├── capstone_project_1_mali.pkl # Trained ML pipeline (scaler + RFE + XGB) └── README.md # Project overview and documentation ``` --- ## 🧠 Selected Features Used in Model The ML model uses the following 10 features: | Feature Name | Description | |--------------------------|-------------| | `OOccupier` | If the person is the main house owner (1 = Yes, 0 = No) | | `OldFA` | Number of old health program visits (numeric) | | `ageyear_2014` | Age of the child in years as of 2014 | | `vaccine_card_available` | Whether a vaccine card was shown (1 = Yes, 0 = No) | | `child_weighing_type_` | Whether the child was weighed in a clinic (1 = Yes, 0 = No) | | `Changed_water_source` | Whether the water source was changed recently (1 = Yes, 0 = No) | | `A …