Name | Loss function (L) | Regularizer (R) |
---|---|---|
AIC/BIC | ∥y−〈ω,x〉∥_{2} | ∥ω∥_{0} |
Lasso | ∥y−〈ω,x〉∥_{2} | ∥ω∥_{1} |
Elastic Net | ∥y−〈ω,x〉∥_{2} | \(\| \omega \|^{2}_{2}\) + ∥ω∥_{1} |
Regularized Least Absolute | ||
Deviations Regression | ∥y−〈ω,x〉∥_{1} | ∥ω∥_{1} |
Classic SVM | max(0,1−y〈ω,x〉)^{a} | \( \frac {1}{2} \| \omega \|^{2}_{2}\) |
ℓ _{1}-SVM | max(0,1−y〈ω,x〉)^{a} | \( \frac {1}{2} \| \omega \|_{1}\) |
Logistic Regression | log(1+exp(−y〈ω,x〉)) | \( \frac {1}{2} \| \omega \|_{1}\) |