Anzeige
Erschienen in:
2011 | OriginalPaper | Buchkapitel
Decimal-Integer-Coded Genetic Algorithm for Trimmed Estimator of the Multiple Linear Errors in Variables Model
verfasst von : Fuchang Wang, Huirong Cao, Xiaoshi Qian
Erschienen in: Information Computing and Applications
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
The multiple linear errors-in-variables model is frequently used in science and engineering for model fitting tasks. When sample data is contaminated by outliers, the orthogonal least squares estimator isn’t robust. To obtain robust estimators, orthogonal least trimmed absolute deviation (OLTAD) estimator based on the subset of
h
cases(out of
n
) is proposed. However, the OLTAD estimator is NP-hard to compute. So, an new decimal-decimal-integer-coded genetic algorithm(DICGA) for OLTAD estimator is presented. We show that the OLTAD estimator has the high breakdown point and appropriate properties. Computational experiments of the OLTAD estimator of multiple linear EIV model on synthetic data is provided and the results indicate that the DICGA performs well in identifying groups of high leverage outliers in reasonable computational time and can obtain smaller objective function fitness.