Sciweavers

DATAMINE
2006
127views more  DATAMINE 2006»
13 years 11 months ago
Computing LTS Regression for Large Data Sets
Least trimmed squares (LTS) regression is based on the subset of h cases (out of n) whose least squares t possesses the smallest sum of squared residuals. The coverage h may be se...
Peter Rousseeuw, Katrien van Driessen