Abstract
A straightforward algorithm is proposed for L1-norm minimisation. The proposed algorithm is based on grey wolf optimisation which is a meta-heuristic method. Although L1-norm is an efficient tool for robust estimation and outlier detection, the complexity of its implementation made it less useful than L2-norm since after formulation of the L1-norm minimisation for a certain problem one must solve a linear programming problem by a search method while here we only need to set the corresponding L1-norm target function. Two geodetic examples approve the efficiency of the proposed approach.
Acknowledgements
I appreciate Dr. Collier and anonymous reviewers for their valuable comments. This study was supported by Golestan University, Iran (Gran Nos. 1578). I am also grateful for this.
Additional information
Notes on contributors
Vahid Mahboub
Vahid Mahboub graduated from the University of Tehran with PhD degree in geodesy. He is currently the Faculty member of Golestan University, Iran. His interests are developing mathematical theories for applications in satellite geodesy and engineering.