Drugs Sale Forecasting Based on SVR Integrated Promotion Factors

Abstract

In order to improve the inventory management level of pharmaceutical chain enterprise, drug sales forecasting is needed. Using machine learning SVR model, we forecast the drugs sale amount with a relatively high accuracy. Suppose the sales amount is influenced by promotion strategy, we integrated the promotion factors into the SVR model. Experimental result of drug sales prediction on the real sales data of a large chain drug company S shows that the SVR algorithm integrated with promotion factors get the accuracy rate of 91% and the algorithm can greatly improve the drug sales prediction results compared with traditional time series model.

The Prediction of Douyin Live Sales based on Neural Network Algorithms

“…Zhan et al [11] developed a Bayesian linear regression method to predict product sales and found that this method performed well in terms of MAE and RMSE. Liu et al [12] designed a support vector regression model to predict drug sales and improve inventory management, achieving an accuracy rate of 91% for this method.…”