Hybrid Gradient Descent Grey Wolf Optimizer for Optimal Feature Selection

Pseudocode 3

Partial derivative binary version.

1. Function partial derivative

2. Pass in: wolf_position, fitness_function

3. Set partial_derivatives as the vector as zeros

4. For each feature of the wolf_position

 a. Set new_position as Call not if wolf_position feature

 b. Set the new_fitness value as Call fitness_function for new_position

 c. Set partial_derivative index as the difference of new_fitness and fitness of wolf_position

5. End for

6. Pass out: partial derivatives

7. End function