Ping Ji

Multinomial Logistic Regression for Interpretable Multiclass Decision Systems: A Systematic Review and Integrated Research Framework

Multinomial Logistic Regression (MLR) remains one of the most widely used interpretable models for multiclass classification, risk prediction, and discrete decision analysis. Its continued relevance does not reflect the novelty of the classical model, but the new demands placed on it by high-dimensional, noisy, imbalan...