overfitting in Greek

(statistics, machine learning) The production of a statistical model that, due to an excess of parameters relative to the sample size, fits the sample data extremely well but fails to fit new data.