Regularized Regression Components

The TemporalMixtureModels.jl package includes support for regularized regression components, specifically Ridge Regression and Lasso Regression. These components are useful when dealing with multicollinearity or when you want to perform variable selection in your regression models. The regularized regression components extend the basic polynomial regression model by adding a penalty term to the loss function.

TemporalMixtureModels.RidgeRegression — Type

RidgeRegression(degree::Int, λ::Real)

Polynomial regression component model with L2 regularization penalization (Ridge regression).

Arguments

degree: Polynomial degree (e.g., 2 for quadratic)
λ: Regularization strength (non-negative)

source

TemporalMixtureModels.LassoRegression — Type

LassoRegression(degree::Int, λ::Real)

Polynomial regression component model with L1 regularization penalization (Lasso regression).

Arguments

degree: Polynomial degree (e.g., 2 for quadratic)
λ: Regularization strength (non-negative)

source