= Truncated singular value decomposition
= Spectral cutoff regularization
{synonym}
For a <singular value decomposition> $A v_j=\sigma_j u_j$, truncated SVD with cutoff $\tau>0$ estimates the inverse action by $x_\tau=\sum_{\sigma_j\geq\tau}\sigma_j^{-1}\langle b,u_j\rangle v_j$. Discarding small singular values limits amplification of data errors; this is a <spectral regularization method>.
Truncated singular value decomposition retains only singular components above a threshold, or equivalently only a finite leading set of singular vectors.
Back to article page
