multicollinearity

Tanımlar (İngilizce)

1. isim A phenomenon in which two or more predictor variables in a multiple regression model are highly correlated, so that the coefficient estimates may change erratically in response to small changes in the model or data.