Hybrid Gradient Descent Grey Wolf Optimizer for Optimal Feature Selection

Pseudocode 4

Implementation 1.

1. Function mutation with partial derivative

2. Pass in: wolf_pos, derivatives, probability

3. Set probability vector as scaled derivative vector multiplied by probability

4. Set selected features as probability vector compared against a generated random number vector

5. Set new_pos as Call XOR of wolf_pos and selected_features

6. Pass out: new_pos