Research Article
Artificial Intelligence-Based Digital Image Steganalysis
Algorithm 2
-nondominated sorting genetic algorithm III-based DCNN model.
| elect randomly solutions from the elite | | | for all do | | | decode as initial parameters of DCNN | | | for to do | | | compute DCNN with random initial parameters in | | | if then | | | | | | end if | | | end for | | | if then | | | | | | end if | | | for to do | | | select randomly an | | | if then | | | | | | else | | | | | | end if | | | if then | | | | | | end if | | | end for | | | end for | | | if then | | | select solutions computed using NSGA-III | | | end if |
|