摘要
提出一种新的基于关键点的时间序列分段拟合算法.通过一次扫描数据,该算法依次利用三个连续数据形成的夹角和非单调序列中的极值点,选择反映序列趋势变化的关键点,实现时间序列的线性拟合的同时剔除了噪音干扰,能精确定位单调序列中的突变转折点,发现序列中的尖峰状态.实验结果表明该算法具有良好的分段拟合性能.
A novel segmentation algorithm based on key points was presented. Three continuous data points were first selected in turn when the data points in time series were scanned. According to the angle formed by these three data and the extreme value in monotone sequence, the key points reflecting the sequence＇s changing feature were then recorded. With these key points, the original time series can be fitted linearly while some small noises are attenuated, and peak subsequences and jump points can be found more accurately. Theoretical analysis and experiment results show the efficiency of new method.
基金
国家重点基础研究发展(973)计划(2006CB705800)
国家自然科学基金(10672159)资助
关键词
时间序列
中线长度
线性拟合
关键点
time series
midline length
linear fitting
key points
