基于复杂网络的时间序列转换算法

Time Series Transformation Algorithms Based on Complex Networks

引入最大值分裂算法和凸包发现算法,分别对时间序列转换为复杂网络的经典可视图算法进行改进,并将经典可视图算法及其2种改进算法应用于分形布朗运动时间序列中.实验结果表明,2种改进算法能明显降低时间复杂度.

To improve classic visibility algorithm of transformation from time series to complex networks,this paper proposes maximum value split algorithm and convex hull algorithm and applies them in the fractal Brown time series respectively.Experimental results show that improved algorithm can reduce time complexity remarkably.