Data Science and Machine learning: A survey on retail sales forecasting and prediction in Rwanda Supermarkets

This article focuses on how supermarkets choose to stock products at different times of the year, in most supermarkets there is no real formal system in place it is just a case of the person in charge of stock making their own decisions based on their past experiences and what they know
and have observed from their clientele. So, most of the estimations are done by people. This is where the proposed research comes in. Instead of solely depending on people making their own estimations we can now utilize trusted calculations and analysis of the previous sales in a supermarket using python libraries and built in functions to make accurate predictions of which products will sell more than others and by what margin.The article aims to find and define the
spending patterns of consumers to help shop owners stock products more efficiently and reduce on stock waste. This article also aims to see if the future purchases of consumers can really be predicted. The recommendation is that this prediction tool can be helpful to local supermarkets to possibly increase sales and cut down on products expiring on shelves. This will also help local supermarkets to better plan their financial years.