The human intelligence lies in the algorithm, the nature of algorithms lies in the classification, and the classification is equal to outlier detection. This paper is based on its past unpublished edition (2009), which discussed an application of (pe) algorithm in outlier detection for time series data. The pe algorithm, also named as RDD algorithm, is originated from the study on general AI. United with it, designed modules can be used to realize kinds of tasks. A primary framework concerned with the mind through pe algorithm has been constructed in prior works. In this concise paper, we neglect background and minor description of the early edition, and directly discuss the main contents include longest k-turn subsequence problem, curve type outliers, futural directions and related comments.
View on arXiv