Outlier detection, also referred as anomaly detection, always be considered as a primary step in data analysis. This paper provides a synthesis of some statistical-based methods and their potential applications for detecting outlier in bridge monitoring time series data. The results indicate the combination of InterQuartile Range and Mahalanobis Distance is the most appropriate and effective method to identify abnormal points in GPS time series in terms of three-dimensional coordinates data from bridge health monitoring systems.

Full text: IJEIR_2460_FINAL.pdf