| 摘要: |
| 依托欧拉距离,使用杂合距离算法改进Minkowski距离公式,使得最近邻算法能够针对不同实际需要计算两事例距离,适用到属性是混合型的情形,也能避免时序列中出现的错误计算问题。 |
| 关键词: 邻算法 欧式距离 Minkowski 距离 |
| DOI： |
| 投稿时间：2010-09-16修订日期：2010-10-11 |
| 基金项目:广西自然科学基金项目(桂科自0899018),广西教育厅科研项目(200808MS062)资助。 |
|
| Improved kNN Algorithm Based on Euclidean Distance |
|
LIU Xing-yi, WEI Xiao-ling
|
| (Qinzhou University, Qinzhou, Guangxi, 535000, China) |
| Abstract: |
| Based on Euclidean distance,a hybrid method was employed by making k Nearest Neighbor (kNN) algorithm available to calculate the distance between two instances. The proposed method can be applied for the case with all kinds of data and avoid the mistake computation in the time series data. |
| Key words: nearest neighbor algorithm Euclidean distance Minkowski distance |
