galts: Genetic algorithms and C-steps based LTS (Least Trimmed Squares)
estimation
This package includes the ga.lts function that estimates
LTS (Least Trimmed Squares) parameters using genetic algorithms
and C-steps. ga.lts() constructs a genetic algorithm to form a
basic subset and iterates C-steps as defined in Rousseeuw and
van-Driessen (2006) to calculate the cost value of the LTS
criterion. OLS(Ordinary Least Squares) regression is known to
be sensitive to outliers. A single outlying observation can
change the values of estimated parameters. LTS is a resistant
estimator even the number of outliers is up to half of the
data. This package is for estimating the LTS parameters with
lower bias and variance in a reasonable time. Version 1.3
included the function medmad for fast outlier detection in
linear regression.
| Version: |
1.3 |
| Depends: |
genalg, DEoptim |
| Published: |
2013-02-07 |
| Author: |
Mehmet Hakan Satman |
| Maintainer: |
Mehmet Hakan Satman <mhsatman at istanbul.edu.tr> |
| License: |
GPL-2 | GPL-3 [expanded from: GPL] |
| NeedsCompilation: |
no |
| CRAN checks: |
galts results |
Downloads: