Locality-Based Visual Outlier Detection Algorithm for Time Series

Abstract

Physiological theories indicate that the deepest impression for time series data with respect to the human visual system is its extreme value. Based on this principle, by researching the strategies of extreme-point-based hierarchy segmentation, the hierarchysegmentation-based data extraction method for time series, and the ideas of locality outlier, a novel outlier detection model and method for time series are proposed. The presented algorithm intuitively labels an outlier factor to each subsequence in time series such that the visual outlier detection gets relatively direct. The experimental results demonstrate the average advantage of the developed method over the compared methods and the efficient data reduction capability for time series, which indicates the promising performance of the proposed method and its practical application value.

Anomaly-Aware Traffic Prediction Based on Automated Conditional Information Fusion

“…One group of anomaly detection methods are based on distribution, distance or density. This group includes traditional distribution based methods like dynamic Poisson distribution based detection [19], Chi-square test [20], and modern machine learning methods such as local outlier factor [21], [22], one class support vector machine [23] and iForest [24]. However, distance/density-based methods require the time series should be transformed into points in a multidimensional space.…”

Outlier Detection for Improving Data Robust by ODAD Clustering Technique

“…Outliers existence can be detected in conventional data sets or big data [4]. According to Jiawei Han, Outliers usually get generated either by measurement or execution error or consequence of inherent data variability [5]. Data mining algorithms aims at minimizing or eliminating the influence or presence of outliers.…”