Predictive modelling of sales return management with considering product expiry using machine learning model

Abstract

"Past period of 5 years was quite challenging for Sri Lankan businesses. With the extreme events happened in past period many of domestic businesses are in a critical situation and seeking. many initiatives to overcome these obstacles. Sales Force Automation (SFA) is critical. component in modern retail sales operation in many industry verticals. SFA is mainly involves. in development of automating sales tasks, workflows to reduce manual repetitive efforts on the retail sales execution. There are many organizations who use this type of systems in many. industries such as Fast-Moving Consumer Goods (FMCG), pharmaceuticals, electronics, hardware items, oils and fats, alcohol etc. Industries such as FMCG, oils & fats and pharmaceuticals have items which are longer lifespan as well as shorter lifespan. Managing shorter lifespan items are curtailed in these industries because they must maintain. manufacturing in a most effective way since expiring items in the market will make a significant impact to the business. Currently this process is not well managed in many. organizations, and it create a huge loss of income throughout the process. In this paper it seeks to find a suitable method to overcome these issues using the knowledge in machine learning and deep learning processes and techniques. The proposing solution is to determine to find out whether there are any excess stocks laying within the retail shops than the expected sales forecast and if so transfer them to a suitable retailer for sales prior to the expiry of the product rather than taking expired returns afterwords. The previous sales values are used to determine the sales forecasting and predict future sales on the town level. Proposing a web application as part of the solutions which can also be use in mobile devices. "