Smart Sales Prediction of Pharmaceutical Products

Abstract

Sales prediction is a predominant area in business intelligence. It plays a significant role in supply chain management. Proper sales prediction is essential for pharma companies. Before launching the pharmaceutical product to the market, the producer should predict the sales of the product in that particular area. In case of missing data or lack of adequate data makes the prediction more complex. To predict sales accurately, we use different machine learning algorithms. We can find complicated patterns in the sales dynamics including various risk variables in detailed study and analysed comprehensible predictive models to improve future sales predictions. Building a model based on historical data to forecasting sales of medicines, which can be applicable to new drugs which are licensed and released for sales. A way to show the effectiveness of the forecasting sales in drugs, taking the factors influencing, revealing the reviews of the existing solutions and analysing specific areas. We have tested with 5 different machine learning algorithms with the pharmaceutical product dataset and reached to a best algorithm i.e. linear regression. Its performance, Mean absolute percentage error (MAPE) is 19.07% and is better than other performing model. Hence our experiment shows the linear regression model is the best model for predicting pharmaceutical product sales.