CD Skripsi
Koefisien Determinasi Pada Model Regresi Robust
This final project discusses the coefficient of determination R2 as one measure
that can be used to find out the model in accordance with the data outliers.
Coefficient R2 is obtained by the method of ordinary least squares (OLS), but
the OLS method is sensitive to outliers. Then a method robust against outlier
called the robust regression method least trimmed squares (LTS) is used. The
results show, based on the value of determination coefficient and standard error,
that the LTS method is better than the OLS method.
Keywords: Robust regression, R2-coefficient, least trimmed squares, outlier
detection
Tidak tersedia versi lain