摘要
本文根据旅游信息服务的特点,在多元统计分析中原有的马氏距离判别法的基础上,提出了一种加权的马氏距离判别法,并运用主成分分析思想,得到了确定权值的方法.该方法运用于网络旅游信息服务智能推荐系统,通过78位注册用户的实际数据,与传统的马氏距离判别法和贝叶斯判别法进行比较,证实了加权马氏距离判别法是十分有效的.
Based on the Mahalanobis distance method in multivariate statistical analysis, the paper presents a new method for discriminant analysis in tourism service——Weighted Mahalanobis Distance method（WMD） .In the new approach, the weight can be accurately determined with the Principal Component Analysis （PCA）scheme. Moreover, WMD is compared with two traditional methods Mahalanobis Distance method and Bayesian Discriminant method -- to prove its effectiveness in diseriminant analysis, by analyzing 78 people＇s data from the Tourism Information Recommender System（TIRS）.
出处
《经济数学》
2007年第2期185-188,共4页
Journal of Quantitative Economics
关键词
马氏距离
加权马氏距离
主成分
智能推荐
Mahalanobis distance method, weighted Mahalanobis distance method, PCA, intelligent recommendation
