Whether it is parameter estimation, data smoothing or statistical evaluations – least squares is the workhorse of geodetic data analysis. But it is not an inevitability.
In this tutorial we will survey modern open-source algorithms for optimization and use them to formulate and solve estimation problems beyond the capability of least squares. Python notebooks help us in practically integrating robustness guarantees, equalities, and inequalities into estimation problems. Concrete application examples from engineering geodesy and a discussion of similarities and differences to classical adjustment theory complement the practical and interactive parts of this tutorial.
