1
$\begingroup$

\begin{equation} \label{eqs.1} \tilde{y}=\sum_{m=1}^{M}\alpha_{m}\kappa(x^{(m)},\tilde{x}) (1) \end{equation} where $\kappa(x^{(m)},\tilde{x})$defines the similarity between data and can be chosen beforehand.The weighting$\alpha=(\alpha_{1},...,\alpha_{M})^{T}$,for instance,can be determined by minimizing the least-square loss function \begin{equation} \label{eqs.2} L(\alpha)=\sum_{m=1}^{M}(\tilde{y}^{(m)}-y^{(m)})^{2}+\chi\sum_{m=1}^{M}\alpha_{m}^{2} (2) \end{equation} The combination of (1) and (2) is a kernel ridge regression. How to get the solution and prediction of nuclear ridge regression?

$\endgroup$
1
  • 4
    $\begingroup$ I’m voting to close this question because its off topic. It would be more appropriate for the Cross Validated stackexchange site $\endgroup$ – forky40 Mar 18 at 18:59

Browse other questions tagged or ask your own question.