Forecasting Walmart Store Weekly Sales Using Vector Autoregression with Exogenous Factors

This paper, a model for predicting the sales of Walmart is demonstrated with the VARMAX technique (Vector Autoregression with Exogenous Factors). By incorporating sales data along with variables like holidays as well as economic indicators such as temperature fluctuations, fuel costs, CPI (Consumer Price Index) and unemployment rates, the Walmarts weekly sales figures can be precisely forecasted by this model. Impressively, the performance metrics of thi model achieved an RMSE of 18, 961 and a MAPE of 4%, outperforming forecasting models such as ARIMA. Considering these influences, the model successfully recognized seasonal sales growth and considered economic changes that may affect consumer habit.