Study on Isolated Singularity Data Mining Algorithm based on Sliding Window
Wen-Gang Che · Jiangxi kexue · 2011
Isolated Singularities are series points which are remarkably different from others in the Finance time series.This paper presents an Isolated Singularity mining algorithm based on slip window.The algorithm detects the isolated singularities in the finance time series using local outlier factor detecting method,then deduces the trend of the Isolated Singularity using the moving average model,by doing this we can make out the compact of the isolated singularity on trend of finance time series.By analyzing securities information and the time series composed by Shanghai stock exchange composite index,argues the rationality and validity of the algorithm.