摘要
针对药品销售的高度随机性和非线性,应用RBF神经网络建立药品销售预测模型,并将其与多元线性回归、支持向量机和BP神经网络预测模型的预测效果进行比较。研究结果表明,利用RBF神经网络预测模型预测药品销售的误差小,可以更准确地对药品销售进行预测。
The drug sales play a very important role in safeguarding the human health and life.In view of high stochastic and nonlinear of drug sales,RBF neural network is used to establish the drug sales forecasting model,through the multiple linear regression,the support vector machine and BP neural network prediction models are compared,it is higher precision,smaller error of the RBF neural network prediction,it provides a new practical and effective method for the prediction of drug sales.
出处
《长江大学学报(自科版)(上旬)》
CAS
2013年第7期65-67,89,共4页
JOURNAL OF YANGTZE UNIVERSITY (NATURAL SCIENCE EDITION) SCI & ENG
基金
安徽省教育厅自然科学基金资助项目(KJ2012Z262)