MODERN METHODS OF REGRESSION ANALYSIS AND THEIR APPLICATION IN STATISTICAL MODELING

Abstract

This article explores modern methods of regression analysis applied in statistical modeling. It focuses on both classical approaches such as the Ordinary Least Squares (OLS) method and robust regression techniques that address outliers and violations of standard model assumptions. The study examines the strengths and limitations of various models, comparing parameter estimation accuracy and model stability. Model selection criteria, including AIC and BIC, are also discussed. The empirical section includes numerical simulations and graphical illustrations demonstrating the effectiveness of different approaches. The findings highlight the importance of selecting an appropriate regression method to ensure the reliability and interpretability of statistical results, particularly in advanced academic research at the doctoral level.