Outlier Detection and Trend Detection: Two Sides of the Same Coin

摘要: Outlier detection is commonly defined as the process of finding unusual, rare observations in a large data set, without prior knowledge which objects to look for. Trend task some unexpected change quantity, such occurrence certain topics textual stream. Many established outlier methods are designed search for low-density static set vectors Euclidean space. For trend detection, high volume events interest and constantly changing. These two problems appear be very different at first. However, they also have obvious similarities. example, trends outliers likewise supposed occurrences. In this paper, we discuss close relationship these tasks. We call action investigate further, carry over insights, ideas, algorithms from one domain other.