| CRM中序列模式分析与神经网络结合的应用研究 |
| |
| 引用本文： | 杨欢耸.CRM中序列模式分析与神经网络结合的应用研究[J].浙江大学学报(理学版),2006,33(2):174-177,181. |
| |
| 作者姓名： | 杨欢耸 |
| |
| 作者单位： | 杭州师范学院,浙江,杭州,310018 |
| |
| 摘 要： | 用序列模式挖掘方法进行商品销售的预测,其准确率受单个算法本身限制,如果先用序列模式挖掘对原始数据进行过滤,然后再利用神经网络的自适应性,通过学习和训练,找出客户特征与购买某种商品意向的内在联系,再利用神经网络的泛化功能对未经训练客户购买意向进行预测,其预测的正确率非常高.
|
| 关 键 词： | 序列 挖掘 预测 数据 神经网络 |
| 文章编号： | 1008-9497（2006）02-174-04 |
| 收稿时间： | 2005-02-28 |
| 修稿时间： | 2005-02-28 |
Research of application of combination of sequence pattern analysis and neural network in CRM |
| |
| YANG Huan-song.Research of application of combination of sequence pattern analysis and neural network in CRM[J].Journal of Zhejiang University(Sciences Edition),2006,33(2):174-177,181. |
| |
| Authors: | YANG Huan-song |
| |
| Institution: | Hangzhou Teachers College, HangZhou 310018, China |
| |
| Abstract: | The veracity of forecast of products sale using sequence pattern detecting is restricted by single algorithm, If filtration is done to the original data by means of sequence pattern tapping; and after being studied and trained, the inherent relation between customer character and the trend of purchasing certain product is find out by using the self-adjustability of neural network; and then forecast for purchase trend of un trained customers is done by using extending function of neural network, a very high veracity of forecast can be obtained. |
| |
| Keywords: | sequences detects forecast data nerval network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《浙江大学学报(理学版)》浏览原始摘要信息 |
| 点击此处可从《浙江大学学报(理学版)》下载免费的PDF全文 |
|
