Big Mart Sales Prediction using Machine Learning Algorithms

This paper discusses the analysis and predictive modeling of sales data for a big mart. The data, sourced from Kaggle, contains product details and outlet information. This will assist organizations in inventory management, profit maximization, supply-chain management, sales forecasting etc. Solution methodology includes data pre-processing, feature engineering, univariate and bivariate analysis. Several machine-learning models, including Linear Regression, Lasso, Ridge Regression, Random Forest, and XGBoost, were evaluated for their predictive accuracy using RMSE as a metric. The study concludes that Random Forest and XGBoost provide the most accurate sales predictions.