摘要
提出采用符号时间序列分析方法获得时间序列的大数据特征.给出对十进制时间序列实施二进制符号化的原理和步骤.对某股市指数变化曲线进行符号时间序列分析,表明多方较之空方占有上风,股指变化过程中同时含有确定性因素和随机性因素的作用.对某四缸柴油机机身振动曲线进行符号时间序列分析,表明柴油机机身振动整体上具有随机性事件属性.基于符号信息的研究结果说明,符号时间序列分析方法能够获得蕴藏在时间序列中的某些大数据特征.
In order to obtain the characteristics of big data in time series, a method for symbolic time series analysis （STSA） is proposed. The principle and procedure to carry out binary symbol for decimal time series are given. The STSA for a stock market index curve indicates that the bulls are stronger than the bears meanwhile there are both the deterministic and stochastic factors acting in the process of stock index changes. The STSA for a four cylinder diesel engine body vibration curve indicates that there is a random event attribute about the diesel engine body vibration as the whole.The result based on the symbol information shows that some characteristics of big data in time series could be obtained by STSA.
出处
《南京工程学院学报（自然科学版）》
2016年第3期30-33,共4页
Journal of Nanjing Institute of Technology(Natural Science Edition)
基金
江苏省自然科学基金项目(BK20130746)
关键词
信息工程
符号时间序列分析
大数据
特征
information engineering
symbolic time series analysis
big data
characteristics
