Smart Poverty Forecast

A machine learning-based application for predicting poverty levels using historical data.

Project Structure

smart-poverty-forecast/
├── backend/          # FastAPI backend server
│   ├── main.py       # API endpoints
│   ├── *.pkl         # Trained ML models
│   └── requirements.txt
├── frontend/         # Streamlit frontend application
│   ├── app.py        # Web interface
│   └── requirements.txt
└── README.md

Setup Instructions

Prerequisites

• Python 3.7+ installed
• pip package manager

Installation

1. Install Backend Dependencies:

   cd backend
   pip install -r requirements.txt

2. Install Frontend Dependencies:

   cd frontend
   pip install -r requirements.txt

Running the Project

Option 1: Using Batch Scripts (Windows)

• Run start_backend.bat to start the backend server
• Run start_frontend.bat to start the frontend application (in a separate terminal)

Option 2: Manual Start

Terminal 1 - Backend:

cd backend
uvicorn main:app --reload

Backend will be available at: http://127.0.0.1:8000

Terminal 2 - Frontend:

cd frontend
streamlit run app.py

Frontend will be available at: http://localhost:8501

API Endpoints

• GET / - Health check endpoint
• POST /predict - Predict poverty levels
  • Parameters:
    • income: Current median household income
    • poverty_last_year: Poverty percentage last year
    • income_last_year: Income from last year
    • avg_3yr: 3-year poverty average

Usage

1. Start the backend server first
2. Start the frontend application
3. Open your browser to the Streamlit URL (usually http://localhost:8501)
4. Enter the required values and click "Predict Poverty Levels"

Notes

• The backend server must be running before using the frontend
• Models are loaded automatically when the backend starts
• Use --reload flag for development (automatic restart on code changes)