Hildreth–Lu estimation

Hildreth–Lu estimation, named for Clifford Hildreth and John Y. Lu,[1] is a method for adjusting a linear model in response to the presence of serial correlation in the error term. It is an iterative procedure related to the Cochrane–Orcutt estimation.

The idea is to repeatedly apply ordinary least squares to

for different values of between −1 and 1. From all these auxiliary regressions, one selects the pair (α, β) that yields the smallest residual sum of squares.

See also

edit

References

edit
  1. ^ Hildreth, C.; Lu, J. Y. (November 1960). "Demand Relations with Autocorrelated Disturbances". Technical Bulletin. 276. Michigan State University Agricultural Experiment Station.

Further reading

edit