Nieklasyczne metody regresji a problem odporności
Non-classical regression methods vs. robustness

Summary/Abstract: The paper presents an important problem of robustness in regression. Various aspects of the approach to this problem are discussed, but the paper focuses on the sensitivity of the model to outliers, noise of the values of dependent variable and to the presence of insignificant variables. The study assesses the robustness of the following methods: PPR, POLYMARS, RANDOM FORESTS, MART and ridge regression.

• Details
• Contents