Building a Regression Model to Predict Sales of a Supermarket
Big Mart, a fictional retailer of supermarkets and grocery stores, has collected sales data for 1559 different products across 10 outlets in the year 2013. Using the data, Big Mart will try to understand the key aspects of its products and then try to identify key factors that determine sales of products.
The aim is to build a predictive model to estimate the sales of products.
The Ordinary Least Squares algorithm will be used here to build the predictive model. Using backward elimination method, the model's fit will be improved.
Methodology: After loading the data and looking at the data types & columns, the data was cleaned (handling of missing values & invalid values). Then, the data was pre-processed to split into 80:20 for model training & predictions. Fianlly, an initial OLS model was built and fine-tuned using the backward elimination method.
About
Building a Regression Model to Predict Sales of a Supermarket