
DataStructured or unstructed dataIID(Independent identically distributed) data or non-IID dataModelTraining(algorithm & inference)Nonparametaric model

여기서 minimize 하는게 목표y : target bector∅ : design matrixgredient descent method의 하나임(P(n-1) ϕ\_n)/(λ+ϕ\_n^T P(n-1) ϕn ) : constant stepsize\[y_n-ϕ\_n^T

📌 ch 3 . Linear Models for Regression _ Regularization, Shrinkage, Sparseness Ridge Regression > LASSO > Bias-Variance Dilemma > Proximal gradient method(optional) >
CHAP4 Logistic Regression