svt: Singular value thresholding
In riverKangg/NuclearNormClassifier: Nuclear Norm Classifier

Description Usage Arguments Value Author(s) References See Also Examples

View source: R/admm_nmr.R View source: R/fast_admm_nmr.R

This function performs singular value thresholding. Singular value thresholding is used to update y that cannot be represented in closed form. Entering the parameters mu, A (x), B, and Z returns the y after performing singular value thresholding.

1	svt(mu, A_x, B, Z)

`mu`	"mu" is a parameter that must be set when defining a function.
`A_x`	"A_x" is a result obtained by correlation coefficient x with A, a test image data set. This can be made using the coef_img function.
`B`	"B" is an image of a face with occlusion.
`Z`	"Z" can be found using A, B, and y. It can be calculated naturally during algorithm implementation.

svt_out

"svt_out" is the y after performing singular value thresholding.

Jisun Kang

https://ieeexplore.ieee.org/document/7420697

https://arxiv.org/abs/0810.3286

function (mu, A_x, B, Z)
{
    svt_in <- A_x - B + 1/mu * Z
    result_svd <- svd(svt_in)
    d <- pmax(result_svd$d - 1/mu, 0)
    svt_out <- result_svd$u %*% d %*% result_svd$v
    return(svt_out)
  }