A posteriori regularization parameter choice rule for a modified kernel method for a time-fractional inverse diffusion problem.
Songshu LiuLixin FengPublished in: J. Comput. Appl. Math. (2019)
Keyphrases
- regularization parameter
- kernel methods
- image restoration
- cross validation
- support vector
- hyperparameters
- kernel function
- machine learning
- feature space
- reproducing kernel hilbert space
- kernel matrix
- support vector machine
- maximum likelihood
- anisotropic diffusion
- classification accuracy
- total variation
- random sampling
- image processing
- data sets