2 problems
- 0 votes0 replies0 views
Minimal-penalty condition for orthonormal hard thresholding
Consider variable selection with an orthonormal design and Gaussian noise, with denoting model dimension and the noise variance. In this setting, the minimal penal…
- 0 votes0 replies1 view
Eliminating the optimization error in the hard-thresholding bound
Let denote the optimization error appearing in Theorem. Optimization-error conjecture. It should be possible to remove the term from the theorem's bound. The conjectu…