Machine Learning for Revenue Forecasting: A Case Study in Retail business

One of the critical aspects of staying ahead of the competition is to embrace technological advancements and to manage the change successfully. One such advanced technology is Machine Learning (ML), which finds applications in several businesses. We present an application of ML in forecasting the future revenue generation of a retail chain based on previous sales values along with several non-intuitive external factors such as economic performance, consumer price index, unemployment rate, fuel prices, ambient temperature, and price markdowns that may have a causal relationship with revenue generation. The model will provide insights into how the external factors affect the revenue of a retail business. We have employed techniques like Random Forest Regression and Vector Autoregression (VAR). We aim to evaluate the effectiveness of these methods by comparing their MAPE (Mean Absolute Percentage Error) among themselves and with univariate regression techniques such as Auto-Regressive Integrated Moving Average (ARIMA).