Operations-Research-Spektrum

, Volume 8, Issue 3, pp 181–184

# Average linear least squares regression

• H. Späth
OR-Software

## Summary

We describe an algorithm and a corresponding FORTRAN subroutine for finding a regression vectorx withn components such that
$$F(x) = \sum\limits_{i = 1}^s {w_i \left\| {A_i x - b_i } \right\|}$$
is minimized, where ∥.∥ denotes the Euclidean norm,W i >0,A i are design matrices withm i rows andn columns, andb i are observation vectors withm i components (i=1,...,s).

## Keywords

Euclidean Norm Observation Vector Design Matrice FORTRAN Subroutine
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

## Zusammenfassung

Wir beschreiben einen Algorithmus und eine zugehörige FORTRAN-Subroutine zur Auffindung einesn-komponentigen Parametervektorsx derart daß
$$F(x) = \sum\limits_{i = 1}^s {w_i \left\| {A_i x - b_i } \right\|}$$
minimiert wird, wobei ∥.∥ die Euklidische Norm bezeichnet,W i >0,A i Versuchsmatrizen mitm i Zeilen undn Spalten undb i Beobachtungsvektoren der Längem i (i= 1,...,s)sind.

## References

1. 1.
Eckhardt U (1975) On an optimization problem related to minimal surfaces with obstacles. In: Bulirsch R, Oettli W, Stoer J (eds) Optimization and optimal control. Springer, Berlin Heidelberg New YorkGoogle Scholar
2. 2.
Eckhardt U (1980) Weber's problem and Weiszfeld's algorithm in general spaces. Math Progr 18:186–196Google Scholar
3. 3.
Lawson CL, Hanson RL (1974) Solving least squares problems. Prentice Hall, Englewood CliffsGoogle Scholar
4. 4.
Overton ML (1983) A quadratically convergent method for minimizing a sum of Euclidean norms. Math Progr 27:34–63Google Scholar
5. 5.
Späth H (1985) Cluster dissection and analysis, theory, FORTRAN programs, examples. Horwood, ChichesterGoogle Scholar
6. 6.
Voss H, Eckhardt U (1980) Linear convergence of generalized Weiszfeld's method. Computing 25:243–251Google Scholar