In the paper, in order to improve the stability of the deconvolution algorithms based on B-splines, a nonlinear regularization by minimizing the curvature of pressure derivative response, as used in von Schroeter et al.'s algorithm, is appended instead of the linear regularization.
In our method, we use a TV flow-based nonlinear diffusion technique [27], which is the parabolic counterpart to TV regularization [18], instead of TV regularization.
In addition, we applied the Wiener filter to the method in [8], instead of the bilateral-total variation (BTV) regularization to see the effect of the measurement validation only.
We here use quadratic regularization instead of L1 regularization.
In order to avoid overfitting and instead of adding a regularization parameter in the criterion, early stopping is introduced as a regularization method for HK learning, which becomes HKES (Ho Kashyap with early stopping).
The training for selection and classification requires the choice of the regularization parameters for both l1l2 regularization and regularized least squares (RLS) denoted with τ* and λ*, respectively.
Svergun, D. I. Determination of the regularization parameter in indirect-transform methods using perceptual criteria.
In this approach, selection of the regularization map and the regularization parameter is very important.
Under the posterior regularization parameter choice rule, the convergence order of the regularization solution is obtained.
