An Empirical Investigation of the Sales-Promotion Relationship: Modeling Factors Through Linear Regression Analysis

Nowadays, Machine Learning has not only become a stepping stone towards entering a world full of Artificial Intelligence, but has also helped humans to forecast the things which would otherwise be impossible to tell through manual human intervention. Taking this inspiration ahead, a current research study has been conducted to verify how Linear Regression - a Supervised Machine Learning Algorithm represents the relationship between sales and promotion factors. Since the predictions can be done on the Promotion values of the company, depending upon its Sales value. The objective of this research is to identify whether Simple Linear Regression is the best fit algorithm for the case study under consideration and whether the trained model can accurately predict results. The model is provided with the variables that play the most important roles in determining the company's promotional values. The model was trained using Linear Regression in order to give some predictions for the promotion values of the company based on the given sales values. However, it has been observed that high RMSE value and extremely low R2 value are obtained and model is not a perfect fit for the data.