Commodity Price Forecasting System Using SARIMAX Model with Interactive GUI

Abstract

This project presents an intelligent commodity
price forecasting system designed to predict future market
trends using historical pricing data. Leveraging the
Seasonal AutoRegressive Integrated Moving Average with
eXogenous regressors (SARIMAX) model, the system
performs time series analysis to forecast commodity prices
over a five-year horizon. Developed with Python, the
application features a user-friendly graphical interface
built using Tkinter, allowing users to select a commodity,
view historical trends, and visualize forecasted values
through integrated plots. The system reads and
preprocesses time series data from a CSV file, applies the
SARIMAX model to generate monthly forecasts, and
displays both numerical and graphical output to enhance
interpretability. Additional functionality includes RMSEbased
model evaluation and dynamic forecast
visualization using Matplotlib embedded within the GUI.
This tool is particularly useful for researchers, traders,
and policymakers aiming to understand market behavior
and make data-driven economic decisions. Future
enhancements could include multi-model support, web
deployment, and real-time data integration for enhanced
predictive capabilities.