regressiomallin
Regressiomallin is a hypothetical statistical framework that extends standard regression models by introducing a Mallin regularization term to the objective function. The aim is to balance model fit with control of coefficient complexity, allowing simultaneous sparsity and smoothness in the estimated coefficients.
Origin and terminology: The name combines a Latin root for regression with a fictional Mallin penalty, used
Mathematical formulation: For data (X, y), regressiomallin minimizes a loss L(y, Xβ) plus a Mallin penalty P_Mallin(β).
Computation and properties: Optimization can be performed with proximal gradient methods or coordinate descent when P_Mallin
Applications and scope: Regressiomallin is described as suitable for feature selection and prediction in datasets with
See also: regression, regularization, lasso, ridge, elastic net, proximal methods.