| 基于串行分类算法的不平衡时间序列多分类方法 |
| |
| 引用本文： | 陈俐名,黄诗茹,修保新,周鋆.基于串行分类算法的不平衡时间序列多分类方法[J].南京师范大学学报,2019,0(3). |
| |
| 作者姓名： | 陈俐名 黄诗茹 修保新 周鋆 |
| |
| 作者单位： | 国防科技大学信息系统工程重点实验室,湖南 长沙 410073 |
| |
| 摘 要： | 提出了基于串行分类算法的不平衡时间序列多分类方法,并以“上证50指数”15 min交易数据为例,进行了实验检验与结果分析. 结果表明,在多数情况下,串行分类算法比单一算法有更高的准确率、召回率和F1值,可以更有效解决不平衡时间序列多分类问题.
|
| 关 键 词： | 不平衡 时间序列 多分类 串行分类算法 |
Imbalanced Time Series Multi-Classification MethodBased on Two-Steps Algorithm |
| |
| Chen Liming,Huang Shiru,Xiu Baoxin,Zhou Yun.Imbalanced Time Series Multi-Classification MethodBased on Two-Steps Algorithm[J].Journal of Nanjing Nor Univ: Eng and Technol,2019,0(3). |
| |
| Authors: | Chen Liming Huang Shiru Xiu Baoxin Zhou Yun |
| |
| Affiliation: | Science and Technology on Information Systems Engineering Laboratory,National University of Defense Technology,Changsha 410073,China |
| |
| Abstract: | This paper proposes a multi-classification method of imbalanced time series based on two-steps classification algorithm,and takes the 15-minutes trading data of“Shanghai Stock Exchange 50 index”as an example to conduct an experimental test and analysis. The results show that,in most cases,the two-steps classification algorithm has a higher accuracy,recall rate and F1 value than the single algorithm,and that it can solve the problem of multiple classification of unbalanced time series more effectively. |
| |
| Keywords: | imbalance time series multi-classification two-steps algorithm |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《南京师范大学学报》浏览原始摘要信息 |
|
点击此处可从《南京师范大学学报》下载全文 |
|
