Walmart Sales Forecasting

Time Series Analysis for Business Forecasting

This project focuses on analyzing and forecasting weekly sales for a Walmart store using time series analysis techniques. Leveraging historical data from Kaggle's "Walmart Recruiting - Store Sales Forecasting," advanced models were developed to uncover seasonal trends and predict future sales. The analysis provided actionable insights to optimize inventory management, staffing efficiency, and targeted marketing campaigns.



Models and Results

Studied Models
  • SARIMA(0,0,7)(0,1,0)52
  • SARIMA(1,0,0)(0,1,0)52
  • SARIMA(1,1,1)(0,1,0)52
  • SARIMA(0,1,1)(0,1,0)52
Key Results

The SARIMA(1,1,1)(0,1,0)52 model was identified as the best performer, achieving high accuracy in capturing seasonal patterns and trends. It effectively forecasted weekly sales, with most predictions falling within the 95% confidence interval. These results highlight the importance of seasonal differencing and careful parameter selection in time series modeling. The analysis provides Walmart with valuable tools to enhance decision-making in inventory management and marketing strategies.


Final Paper

The PDF is not available for this screen size