Two variants of M split estimation – similarities and differences

Abstract

M split estimation is a novel method developed to process observation sets that include two (or more) observation aggregations. The main objective of the method is to estimate the location parameters of each aggregation without any preliminary assumption concerning the division of the observation set into respective subsets. Up to now, two different variants of M split estimation have been derived. The first and basic variant is the squared M split estimation, which can be derived from the assumption about the normal distribution of observations. The second variant is the absolute M split estimation, which generally refers to the least absolute deviation method. The main objective of the paper is to compare both variants of M split estimation by showing similarities and differences between the methods. The main dissimilarity stems from the different influence functions, making the absolute M split estimation less sensitive to gross errors of moderate magnitude. The empirical analyses presented confirm that conclusion and show that the accuracy of the methods is similar, in general. The absolute M split estimation is more accurate than the squared M split estimation for less accurate observations. In contrast, the squared M split estimation is more accurate when the number of observations in aggregations differs much. Concerning all advantages and disadvantages of M split estimation variants, we recommend using the absolute M split estimation in most geodetic applications.

Keywords

• accuracy
• influence function
• absolute msplit estimation
• squared msplit estimation