This project is based on the Walmart Recruiting – Store Sales Forecasting competition on Kaggle. In this challenge, the task is to predict future weekly sales (in dollars) for combinations of stores and departments using historical retail data from 45 Walmart stores. The dataset includes features such as store characteristics, holiday flags, fuel prices, temperature, and markdown information.
The competition is a time series regression problem where models must forecast weekly sales for the test set. Forecast accuracy is evaluated using the Weighted Mean Absolute Error (WMAE), which places extra weight on holiday weeks to reflect their business importance.
Using time series modeling and regression approaches, my final submission achieved a Kaggle score of 3349.508.
Competition link: https://www.kaggle.com/competitions/walmart-recruiting-store-sales-forecasting