The Outlier Detection Approach for Multivariate Time Series Based on PCA Analysis
Xiaoning Song · Journal of Jiangxi Normal University · 2012
By means of cumulative contribution rate,dimension reduction method based on principal component analysis and principal components of multivariate time series was selected,an efficient multivariate time series outlier detection algorithm was provided based on the k-nearest neighbor local outlier detection algorithm was provide here,.the experimental results show that the algorithm can morely improve the efficiency of multivariate time series outlier detection.