Hybrid Gradient Descent Grey Wolf Optimizer for Optimal Feature Selection

Pseudocode 6

Implementation 3.

1. Function mutation with partial derivative

2. Pass in: wolf_pos, derivatives, probability threshold, a

3. Set weight as 0.4a + 0.1

4. Set the expoilation_probabilities as mapped derivative on the sigmoid curve centred at zero

5. Set the exploration_probalilities as the ratio of features selected to total features for the selected features and 1- this for negative

6. Set selected_features as the sum of weight expoilation_probabilities and 1 – weight exploration_probalilities multiplied by probability_threshold compared against a vector of randomly generated numbers