Abstract
There are two fundamental approaches to determine outliers; these are, the Conventional Outlier Detection Test Procedures (e.g., data-snooping of Baarda, τ-test of Pope etc.) and Robust Estimation Methods.As is known, the Least Squares Estimation (LSE) method is very sensitive to outliers and it spreads the corrupt influence of the outliers upon the good observations. Therefore the Baarda and Pope test methods derived from the LSE are thought of as unsuccessful methods on outlier detection. It is asserted that in case of more than one outlier, these Conventional Outlier Detection Test Procedures become inefficient. In this situation, Robust Estimation Methods are proposedfor application. In this study these possibilities are investigated in trilateration networks structured artificially.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call