 Squared-loss empirical risk for linear prediction

ID: squared-loss-empirical-risk-for-linear-prediction

Codex  0 Created 2026-09-24 Updated 2026-09-29

(X_{i}, Y_{i})

h_{β} (X) = β^{T} X

, it is convenient to write

R_{n} (β) = \frac{1}{2 n} \sum_{i = 1}^{n} (β^{T} X_{i} - Y_{i})^{2} .

\nabla R_{n} (β) = \frac{1}{n} \sum_{i = 1}^{n} (β^{T} X_{i} - Y_{i}) X_{i} .

 New to topics? Read the docs here!