mohammedshuli/maize-price-forecasting-system

SARIMA-based maize price forecasting and decision support system for smallholder farmers in Tanzania. # Maize Price Forecasting System This project builds a practical decision-support system for forecasting maize prices in two Tanzanian regions: Iringa and Mbeya. It combines time-series forecasting with a simple Streamlit dashboard that helps users decide whether to sell or store maize based on expected price movements. ## What this project does - Loads and preprocesses historical maize price data - Detects outliers and handles missing values - Trains forecasting models for each region - Generates price forecasts and confidence intervals - Provides a web dashboard for actionable recommendations ## Project structure - dashboard/app.py — Streamlit user interface - src/ — preprocessing, modeling, validation, and recommendation logic - data/ — raw, processed, and forecast output data - models/ — trained forecasting model artifacts - reports/ — analysis and comparison documents - notebooks/ — exploratory and development notebooks ## Requirements Install dependencies with: ```bash pip install -r requirements.txt ``` ## Run the dashboard Start the application with: ```bash streamlit run dashboard/app.py ``` ## Example workflow 1. Prepare or update the dataset under data/ 2. Train or refresh the forecasting models 3. Generate fresh forecasts 4. Open the dashboard to view recommendations ## Notes - The project is designed for local experimentation and small-scale deployment. - Forecast outputs are stored in data/outputs/. - Model artifacts are stored in models/.